A First Course of

Partial Diﬀerential Equations
in Physical Sciences and Engineering
Marcel B. Finan
Arkansas Tech University
First Draft August 2009
2
Preface
Partial diﬀerential equations are often used to construct models of the most
basic theories underlying physics and engineering. The goal of this book is to
develop the most basic ideas from the theory of partial diﬀerential equations,
and apply them to the simplest models arising from the above mentioned
ﬁelds.
It is not easy to master the theory of partial diﬀerential equations. Unlike
the theory of ordinary diﬀerential equations, which relies on the fundamental
existence and uniqueness theorem, there is no single theorem which is central
to the subject. Instead, there are separate theories used for each of the major
types of partial diﬀerential equations that commonly arise.
It is worth pointing out that the preponderance of diﬀerential equations aris-
ing in applications, in science, in engineering, and within mathematics itself,
are of either ﬁrst or second order, with the latter being by far the most preva-
lent. We will mainly cover these two classes of PDEs.
This book is intended for a ﬁrst course in partial diﬀerential equations at
the advanced undergraduate level for students in engineering and physical
sciences. It is assumed that the student has had the standard three semester
calculus sequence, and a course in ordinary diﬀerential equations.
Marcel B Finan
August 2009
3
4 PREFACE
Contents
Preface 3
Preliminaries 7
1 Some Results of Calculus . . . . . . . . . . . . . . . . . . . . . . . 7
2 Sequences of Functions: Pointwise and Uniform Convergence . . . 14
Review of Some ODEs Results 23
3 The Method of Integrating Factor . . . . . . . . . . . . . . . . . . 23
4 The Method of Separation of Variables for ODEs . . . . . . . . . 28
5 Second Order Linear ODEs . . . . . . . . . . . . . . . . . . . . . 33
Introduction to PDEs 43
6 The Basic Concepts . . . . . . . . . . . . . . . . . . . . . . . . . 43
7 Solutions and Related Topics . . . . . . . . . . . . . . . . . . . . 52
First Order Partial Diﬀerential Equations 65
8 Classiﬁcation of First Order PDEs . . . . . . . . . . . . . . . . . 65
9 The One Dimensional Spatial Transport Equations . . . . . . . . 71
10 The Method of Characteristics . . . . . . . . . . . . . . . . . . . 80
11 The Cauchy Problem for First Order Quasilinear Equations . . . 87
Second Order Linear Partial Diﬀerential Equations 101
12 Second Order PDEs in Two Variables . . . . . . . . . . . . . . . 101
13 Hyperbolic Type: The Wave equation . . . . . . . . . . . . . . . 106
14 Parabolic Type: The Heat Equation in One-Dimensional Space . 114
15 An Introduction to Fourier Series . . . . . . . . . . . . . . . . . 123
16 Fourier Sines Series and Fourier Cosines Series . . . . . . . . . . 134
17 Separation of Variables for PDEs . . . . . . . . . . . . . . . . . 142
5
6 CONTENTS
18 Solutions of the Heat Equation by the Separation of Variables
Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
19 Elliptic Type: Laplace’s Equations in Rectangular Domains . . . 157
20 Laplace’s Equations in Circular Regions . . . . . . . . . . . . . . 168
The Laplace Transform Solutions for PDEs 181
21 Essentials of the Laplace Transform . . . . . . . . . . . . . . . . 181
22 Solving PDEs Using Laplace Transform . . . . . . . . . . . . . . 196
The Fourier Transform Solutions for PDEs 203
23 Complex Version of Fourier Series . . . . . . . . . . . . . . . . . 203
24 Two Dimensional Fourier Transforms . . . . . . . . . . . . . . . 209
25 Applications of Fourier Transforms to PDEs . . . . . . . . . . . 217
Answers and Solutions 225
Preliminaries
In this chapter we include some results from calculus which we will use often
in the study of partial diﬀerential equations. Details and proof of these results
can be found in most calculus books.
1 Some Results of Calculus
The ﬁrst result provides a mean of showing when a function is zero on an
interval.
Theorem 1.1
(a) Suppose that f is continuous on an interval I ⊂ R such that
_
b
a
f(x)dx = 0
for all subintervals [a, b] ⊂ I. Then f(x) = 0 for all x ∈ I.
(b) Suppose that f : [a, b] →R is continuous and non-negative. If
_
b
a
f(x)dx =
0 then f(x) = 0 on [a, b].
(c) Suppose that f : [a, b] → R is continuous such that
_
b
a
f(x)g(x)dx = 0
for all continuous functions g on [a, b]. Then f(x) = 0 on [a, b].
Solution.
(a) Fix a ∈ I. Let x ∈ I. By the Fundamental Theorem of Calculus we have
0 =
d
dx
_
x
a
f(t)dt = f(x).
Since x was arbitrary, we have f(x) = 0 for all x ∈ I.
(b) Suppose the contrary. That is, suppose that x
0
∈ [a, b] such that f(x
0
) >
0. The δ deﬁnition of continuity of f at x
0
guarantees the existence of an
open interval I ⊂ [a, b] centered at x
0
such that f(x) > 0 for all x ∈ I. Hence,
because f(x) ≥ 0 we must have
_
b
a
f(x)dx ≥
_
I
f(x)dx > 0
7
8 PRELIMINARIES
which contradicts our assumption that the integral is zero. We conclude that
f(x) = 0 on [a, b].
(c) This follows from (b) by taking g(x) = f(x)
Remark 1.1
The above theorem remains valid for functions in two variables. For example,
if f(x, y) is deﬁned for x in an interval I and y in an interval J such that
_
b
a
_
d
c
f(x, y)dxdy = 0
for all [a, b] ⊂ J and [c, d] ⊂ I then f(x, y) = 0 over the rectangle I ×J.
Example 1.1
Let f, g : [a, b] →∞ be such that f(x) ≤ g(x) for all x in [a, b]. Show that if
_
b
a
(g(x) −f(x))dx = 0 then f(x) ≡ g(x) on [a, b].
Solution.
Apply part (b) of previous theorem to the function h(x) = g(x) −f(x)
Partial Derivatives
For multivariable functions, there are two common notations for partial
derivatives, and we shall employ them interchangeably. The ﬁrst is the Leib-
nitz notation that employs the symbol ∂ to denote partial derivative. The
second, a more compact notation, is to use subscripts to indicate partial
derivatives. For example, u
t
represents
∂u
∂t
, while u
xx
represents

2
u
∂x
2
, and u
xxt
becomes

3
u

2
x∂t
.
An important formula of diﬀerentiation is the so-called chain rule. If
u = u(x, y) where x = x(s, t) and y = y(s, t) then
∂u
∂s
=
∂u
∂x
∂x
∂s
+
∂u
∂y
∂y
∂s
.
Likewise,
∂u
∂t
=
∂u
∂x
∂x
∂t
+
∂u
∂y
∂y
∂t
.
1 SOME RESULTS OF CALCULUS 9
Example 1.2
Compute the partial derivatives indicated:
(a)

∂y
(y
2
sin xy)
(b)

2
∂x
2
[e
x+y
]
2
Solution.
(a) We have

∂y
(y
2
sin xy) = sin xy

∂y
(y
2
)+y
2 ∂
∂y
(sin xy) = 2y sin xy+xy
2
cos xy.
(b) We have

∂x
[e
x+y
]
2
=

∂x
e
2(x+y)
= 2e
2(x+y)
. Thus,

2
∂x
2
[e
x+y
]
2
=

∂x
2e
2(x+y)
=
4e
2(x+y)
Example 1.3
Suppose u(x, y) = sin (x
2
+ y
2
), where x = te
s
and y = s +t. Find u
s
and u
t
.
Solution.
We have
u
s
=u
x
x
s
+ u
y
y
s
= 2x cos (x
2
+ y
2
)te
s
+ 2y cos (x
2
+ y
2
)
=[2t
2
e
2s
+ 2(s + t)] cos [t
2
e
2s
+ (s + t)
2
]
Likewise,
u
t
=u
x
x
t
+ u
y
y
t
= 2x cos (x
2
+ y
2
)e
s
+ 2y cos (x
2
+ y
2
)
=[2te
2s
+ 2(s + t)] cos [t
2
e
2s
+ (s + t)
2
]
Often we must diﬀerentiate an integral with respect to a parameter which
may appear in the limits of integration, or in the integrand.
Let f(x, t) be a continuous function in the rectangle {a ≤ x ≤ b} ×{c ≤ t ≤
d}. Assume that
∂f
∂t
is continuous on this rectangle. Deﬁne the function
J(t) =
_
b(t)
a(t)
f(x, t)dx
where a(t) and b(t) are continuously diﬀerentiable functions of t such that
a ≤ a(t) ≤ b(t) ≤ b.
Theorem 1.2
dJ
dt
=
d
dt
_
b(t)
a(t)
f(x, t)dx
=f(b(t), t)b

(t) −f(a(t), t)a

(t) +
_
b(t)
a(t)
∂f
∂t
(x, t)dx
10 PRELIMINARIES
Example 1.4
Consider the heat problem
u
t
= ku
xx
−αu, α > 0, k > 0, 0 < x < L, t > 0
with boundary conditions u
x
(0, t) = 0 = u
x
(L, t) and initial condition u(x, 0) =
f(x). Let E(t) =
1
2
_
L
0
u
2
dx.
(a) Show that E

(t) ≤ 0.
(b) Show that E(t) ≤
_
L
0
1
2
|f(x)|
2
dx.
Solution.
(a) We have
dE
dt
=
1
2
_
L
0

∂t
u
2
(x, t)dx
=
_
L
0
u(x, t)u
t
(x, t)dx = k
_
L
0
u(x, t)u
xx
(x, t)dx −α
_
L
0
u
2
(x, t)dx
= ku(x, t)u
x
(x, t)|
L
0
−k
_
L
0
u
2
x
(x, t)dx −α
_
L
0
u
2
(x, t)dx
=−k
_
L
0
u
2
x
(x, t)dx −α
_
L
0
u
2
(x, t)dx ≤ 0
(b) From (a) we conclude that E(t) is a decreasing function of t > 0. Thus,
E(t) ≤ E(0) =
1
2
_
L
0
u
2
(x, 0)dx =
_
L
0
1
2
|f(x)|
2
dx
The Least Upper Bound
A function f : D →R is said to be bounded from above in D if there is a
constant M such that f(x) ≤ M for all x ∈ D. We call M an upper bound
of f. Note that the numbers, M + 1, M + 2, · · · are also upper bounds of
f. The smallest upper bound of f is called the least upper bound or the
supremum. If M is the supremum of f in D we write
M = sup{f(x) : x ∈ D}.
Note that if N is any upper bound of f in D then M ≤ N.
Example 1.5
Find the supremum of f(x) = sin x.
1 SOME RESULTS OF CALCULUS 11
Solution.
The graph of f is bounded between −1 and 1. Thus, sup{f(x) : x ∈ R} = 1
Example 1.6
Find
sup

¸
¸
¸

2
sin
_
x

_
sin
_
t

¸
¸
¸
: x ∈ R, t > 0
_
Solution.
sup

¸
¸
¸

2
sin
_
x

_
sin
_
t

¸
¸
¸
: x ∈ R, t > 0
_
=
2
12 PRELIMINARIES
Practice Problems
Exercise 1.1
Compute the partial derivatives indicated:
(a)

∂x
(y
2
sin xy)
(b)

2
∂x
2
(e
x
2
y
)
(c)

4
∂x∂y
2
∂z
_
z ln
_
x
2
y
__
.
Exercise 1.2
Find all the ﬁrst partial derivatives of the functions:
(a) f(x, y) = x
4
+ 6

y
(b) f(x, y, z) = x
2
y −10y
2
z
3
+ 43x −7 tan (4y)
(c) f(s, t) = t
7
ln (s
2
) +
9
t
3

7

s
4
(d) f(x, y) = cos
_
4
x
_
e
x
2
y−5y
3
(e) f(u, v) =
9u
u
2
+5v
(f) f(x, y, z) =
xsin y
z
2
(g) f(x, y) =
_
x
2
+ ln (5x −3y
2
)
Exercise 1.3
Let f(x, y) = e
3x
cos y. Compute f
x
(0, 2π).
Exercise 1.4
If z = e
x
sin y, x = st
2
, and y = s
2
t, ﬁnd
∂z
∂s
and
∂z
∂t
.
Exercise 1.5
In the equation
∂u
∂x

∂u
∂y
= x −2y
identify the independent variable(s) and the dependent variable.
Exercise 1.6
Let f be an odd function, that is, f(−x) = −f(x) for all x ∈ R. Show that
for all a ∈ R we have
_
a
−a
f(x)dx = 0.
1 SOME RESULTS OF CALCULUS 13
Exercise 1.7
Let f be an even function, that is, f(−x) = f(x) for all x ∈ R. Show that
for all a ∈ R we have
_
a
−a
f(x)dx = 2
_
a
0
f(x)dx.
Exercise 1.8
Use the product rule of derivatives to derive the formula of integration by
parts
_
uv

dx = uv −
_
u

vdx.
Exercise 1.9
Let u

(x, t) =
2
sin
_
x

_
sin
_
t

_
. Find u
tt
and u
xx
.
Exercise 1.10
Let u

(x, t) =
2
sin
_
x

_
sinh
_
t

_
, where
sinh x =
e
x
−e
−x
2
.
Find u
tt
and u
xx
.
Exercise 1.11
Find
sup

¸
¸
¸

2
sinh
_
t

_
sin
_
x

_
¸
¸
¸
¸
: x ∈ R
_
.
Exercise 1.12
Let u
n
(x, t) = 1 +
e
n
2
t
n
sin nx.
(a) Find sup{|u
n
(x, 0) −1| : x ∈ R}.
(b) Find sup{|u
n
(x, t) −1| : x ∈ R}.
14 PRELIMINARIES
2 Sequences of Functions: Pointwise and Uni-
form Convergence
Later in this book we will be constructing solutions to PDEs involving inﬁnite
sums of sines and cosines. These inﬁnite sums or series are called Fourier
series. Fourier series are examples of series of functions. Convergence of
series of functions is deﬁned in terms of convergence of a sequence of func-
tions. In this section we study the two types of convergence of sequences of
functions.
Recall that a sequence of numbers {a
n
}

n=1
is said to converge to a number
L if and only if for every given > 0 there is a positive integer N = N()
such that for all n ≥ N we have|a
n
−L| < .
What is the analogue concept of convergence when the terms of the sequence
are variables? Let D ⊂ R and for each n ∈ N consider a function f
n
: D →R.
Thus, we obtain a sequence of functions {f
n
}

n=1
. For such a sequence, there
are two types of convergenve that we consider in this section: pointwise con-
vergence and uniform convergence.
We say that {f
n
}

n=1
converges pointwise on D to a function f : D →R if
and only if for a given a ∈ D and > 0 there is a positive integer N = N(a, )
such that if n ≥ N then |f
n
(a) −f(a)| < . In symbol, we write
lim
n→∞
f
n
(a) = f(a).
It is important to note that N is a function of both a and .
Example 2.1
Deﬁne f
n
: [0, ∞) → R by f
n
(x) =
nx
1+n
2
x
2
. Show that the sequence {f
n
}

n=1
converges pointwise to the function f(x) = 0 for all x ≥ 0.
Solution.
For all x ≥ 0,
lim
n→∞
f
n
(x) = lim
n→∞
nx
1 + n
2
x
2
= 0
Example 2.2
For each positive integer n let f
n
: (0, ∞) → ∞ be given by f
n
(x) = nx.
Show that {f
n
}

n=1
does not converge pointwise on D.
2 SEQUENCES OF FUNCTIONS: POINTWISE ANDUNIFORMCONVERGENCE15
Solution.
This follows from the fact that lim
n→∞
nx = ∞ for all x ∈ D
As pointed out above, for pointwise convergence, the positive integer N de-
pends on both the given x and . A stronger convergence concept can be
deﬁned where N depends only on .
Let D be a subset of R and let {f
n
}

n=1
be a sequence of functions deﬁned on
D. We say that {f
n
}

n=1
converges uniformly on D to a function f : D →R
if and only if for all > 0 there is a positive integer N = N() such that if
n ≥ N then |f
n
(x) −f(x)| < for all x ∈ D.
This deﬁnition says that the integer N depends only on the given so that
for n ≥ N, the graph of f
n
(x) is bounded above by the graph of f(x) + and
below by the graph of f(x) −.
Example 2.3
For each positive integer n let f
n
: [0, 1] → R be given by f
n
(x) =
x
n
. Show
that {f
n
}

n=1
converges uniformly to the zero function.
Solution.
Let > 0 be given. Let N be a positive integer such that N >
1

. Then for
n ≥ N we have
|f
n
(x) −f(x)| =
|x|
n

1
n

1
N
<
for all x ∈ [0, 1]
Clearly, uniform convergence implies pointwise convergence to the same limit
function. However, the converse is not true in general.
Example 2.4
Deﬁne f
n
: [0, ∞) → R by f
n
(x) =
nx
1+n
2
x
2
. By Example 2.1, this sequence
converges pointwise to f(x) = 0. Let =
1
3
. Show that there is no positive
integer N with the property n ≥ N implies |f
n
(x) −f(x)| < for all x ≥ 0.
Hence, the given sequence does not converge uniformly to f(x).
Solution.
For any positive integer N and for n ≥ N we have
¸
¸
¸
¸
f
n
_
1
n
_
−f
_
1
n

¸
¸
¸
=
1
2
>
16 PRELIMINARIES
Exercise 2.1 below shows a sequence of continuous functions converging point-
wise to a discontinuous function. That is, pointwise convergence does not
preserve the property of continuity. One of the interesting features of uniform
convergence is that it preserves continuity as shown in the next example.
Example 2.5
Suppose that for each n ≥ 1 the function f
n
: D → R is continuous in D.
Suppose that {f
n
}

n=1
converges uniformly to f. Let a ∈ D.
(a) Let > 0 be given. Show that there is a positive integer N such that if
n ≥ N then |f
n
(x) −f(x)| <

3
for all x ∈ D.
(b) Show that there is a δ > 0 such that for all |x −a| < δ we have |f
N
(x) −
f
N
(a)| <

3
.
(c) Using (a) and (b) show that for |x − a| < δ we have |f(x) − f(a)| < .
Hence, f is continuous in D since a was arbitrary. Symbolically we write
lim
x→a
lim
n→∞
f
n
(x) = lim
n→∞
lim
x→a
f
n
(x).
Solution.
(a) This follows from the deﬁnition of uniform convergence.
(b) This follows from the fact that f
N
is continuous at a ∈ D.
(c) For |x −a| < δ we have |f(x) −f(a)| = |f(a) −f
N
(a) +f
N
(a) −f
N
(x) +
f
N
(x)−f(x)| ≤ |f
N
(a)−f(a)|+|f
N
(a)−f
N
(x)|+|f
N
(x)−f(x)| <

3
+

3
+

3
=

Does pointwise convergenvce preserve integration? In real analysis, it is
proven that pointwise convergence does not preserve integrability. That is,
the pointwise limit of a sequence of integrable functions need not be inte-
grable. Even when a sequence of functions converges pointwise, the process
of interchanging limits and integration is not true in general.
Contrary to pointwise convergence, uniform convergence preserves integra-
tion. Moreover, limits and integration can be interchanged. That is, if
{f
n
}

n=1
converges uniformly to f on a closed interval [a, b] then
lim
n→∞
_
b
a
f
n
(x)dx =
_
b
a
lim
n→∞
f
n
(x)dx.
Now, what about diﬀerentiablility? Again, pointwise convergence fails in
general to conserve the diﬀerentiability property. See Exercise 2.1. Does
uniform convergence preserve diﬀerentiability? The answer is still no as
shown in the next example.
2 SEQUENCES OF FUNCTIONS: POINTWISE ANDUNIFORMCONVERGENCE17
Example 2.6
Consider the family of functions f
n
: [−1, 1] given by f
n
(x) =
_
x
2
+
1
n
.
(a) Show that f
n
is diﬀerentiable for each n ≥ 1.
(b) Show that for all x ∈ [−1, 1] we have
|f
n
(x) −f(x)| ≤
1

n
where f(x) = |x|. Hint: Note that
_
x
2
+
1
n
+

x
2

1

n
.
(c) Let > 0 be given. Show that there is a positive integer N such that for
n ≥ N we have
|f
n
(x) −f(x)| < for all x ∈ [−1, 1].
Thus, {f
n
}

n=1
converges uniformly to the non-diﬀerentiable function f(x) =
|x|.
Solution.
(a) f
n
is the composition of two diﬀerentiable functions so it is diﬀerentiable
with derivative
f

n
(x) = x
_
x
2
+
1
n
_

1
2
.
(b) We have
|f
n
(x) −f(x)| =
¸
¸
¸
¸
¸
_
x
2
+
1
n

x
2
¸
¸
¸
¸
¸
=
¸
¸
¸
¸
¸
¸
(
_
x
2
+
1
n

x
2
)(
_
x
2
+
1
n
+

x
2
)
_
x
2
+
1
n
+

x
2
¸
¸
¸
¸
¸
¸
=
1
n
_
x
2
+
1
n
+

x
2

1
n
1

n
=
1

n
(c) Let > 0 be given. Since lim
n→∞
1

n
= 0 we can ﬁnd a positive integer
N such that for all n ≥ N we have
1

n
< . Now the answer to the question
follows from this and part (b)
Even when uniform convergence occurs, the process of interchanging lim-
its and diﬀerentiation may fail as shown in the next example.
18 PRELIMINARIES
Example 2.7
Consider the functions f
n
: R →R deﬁned by f
n
(x) =
sin nx
n
.
(a) Show that {f
n
}

n=1
converges uniformly to the function f(x) = 0.
(b) Note that {f
n
}

n=1
and f are diﬀerentiable functions. Show that
lim
n→∞
f

n
(x) = f

(x) =
_
lim
n→∞
f
n
(x)
_

.
That is, one cannot, in general, interchange limits and derivatives.
Solution.
(a) Let > 0 be given. Let N be a positive integer such that N >
1

. Then
for n ≥ N we have
|f
n
(x) −f(x)| =
¸
¸
¸
¸
sin nx
n
¸
¸
¸
¸

1
n
<
and this is true for all x ∈ R. Hence, {f
n
}

n=1
converges uniformly to the
function f(x) = 0.
(b) We have lim
n→∞
f

n
(π) = lim
n→∞
cos nπ = lim
n→∞
(−1)
n
which does not
converge. However, f

(π) = 0
Pointwise convergence was not enough to preserve diﬀerentiability, and nei-
ther was uniform convergence by itself. Even with uniform convergence the
process of interchanging limits with derivatives is not true in general. How-
ever, if we combine pointwise convergence with uniform convergence we can
indeed preserve diﬀerentiability and also switch the limit process with the
process of diﬀerentiation.
Theorem 2.3
Let {f
n
}

n=1
be a sequence of diﬀerentiable functions on [a, b] that converges
pointwise to some function f deﬁned on [a, b]. If {f

n
}

n=1
converges uniformly
on [a, b] to a function g, then the function f is diﬀerentiable with derivative
equals to g. Thus,
lim
n→∞
f

n
(x) = g(x) = f

(x) =
_
lim
n→∞
f
n
(x)
_

.
Finally, we conclude this section with the following important result that is
useful when a given sequence is bounded.
2 SEQUENCES OF FUNCTIONS: POINTWISE ANDUNIFORMCONVERGENCE19
Theorem 2.4
Consider a sequence f
n
: D →R. Then this sequence converges uniformly to
f : D →R if and only if
lim
n→∞
sup{|f
n
(x) −f(x)| : x ∈ D} = 0.
Example 2.8
Show that the sequence deﬁned by f
n
(x) =
cos x
n
converges uniformly to the
zero function.
Solution.
We have
0 ≤ sup{|
cos x
n
| : x ∈ R} ≤
1
n
.
Now apply the squeeze rule for sequences we ﬁnd that
lim
n→∞
sup{|
cos x
n
| : x ∈ R} = 0
which implies that the given sequence converges uniformly to the zero func-
tion on R
20 PRELIMINARIES
Practice Problems
Exercise 2.1
Deﬁne f
n
: [0, 1] →R by f
n
(x) = x
n
. Deﬁne f : [0, 1] →R by
f(x) =
_
0 if 0 ≤ x < 1
1 if x = 1.
(a) Show that the sequence {f
n
}

n=1
converges pointwise to f.
(b) Show that the sequence {f
n
}

n=1
does not converge uniformly to f. Hint:
Suppose otherwise. Let = 0.5 and get a contradiction by using a point
(0.5)
1
N
< x < 1.
Exercise 2.2
Consider the sequence of functions
f
n
(x) =
nx + x
2
n
2
deﬁned for all x in R. Show that this sequence converges pointwise to a
function f to be determined.
Exercise 2.3
Consider the sequence of functions
f
n
(x) =
sin (nx + 3)

n + 1
deﬁned for all x in R. Show that this sequence converges pointwise to a
function f to be determined.
Exercise 2.4
Consider the sequence of functions deﬁned by f
n
(x) = n
2
x
n
for all 0 ≤ x ≤ 1.
Show that this sequence does not converge pointwise to any function.
Exercise 2.5
Consider the sequence of functions deﬁned by f
n
(x) = (cos x)
n
for all −
π
2

x ≤
π
2
. Show that this sequence converges pointwise to a noncontinuous
function to be determined.
2 SEQUENCES OF FUNCTIONS: POINTWISE ANDUNIFORMCONVERGENCE21
Exercise 2.6
Consider the sequence of functions f
n
(x) = x −
x
n
n
deﬁned on [0, 1).
(a) Does {f
n
}

n=1
converge to some limit function? If so, ﬁnd the limit func-
tion and show whether the convergence is pointwise or uniform.
(b) Does {f

n
}

n=1
converge to some limit function? If so, ﬁnd the limit func-
tion and show whether the convergence is pointwise or uniform.
Exercise 2.7
Let f
n
(x) =
x
n
1+x
n
for x ∈ [0, 2].
(a) Find the pointwise limit f(x) = lim
n→∞
f
n
(x) on [0, 2].
(b) Does f
n
→f uniformly on [0, 2]?
Exercise 2.8
For each n ∈ N deﬁne f
n
: R →R by f
n
(x) =
n+cos x
2n+sin
2
x
.
(a) Show that f
n

1
2
uniformly.
(b) Find lim
n→∞
_
7
2
f
n
(x)dx.
Exercise 2.9
Show that the sequence deﬁned by f
n
(x) = (cos x)
n
does not converge uni-
formly on [−
π
2
,
π
2
].
Exercise 2.10
Let {f
n
}

n=1
be a sequence of functions such that
sup{|f
n
(x)| : 2 ≤ x ≤ 5} ≤
2
n
1 + 4
n
.
(a) Show that this sequence converges uniformly to a function f to be found.
(b) What is the value of the limit lim
n→∞
_
5
2
f
n
(x)dx?
22 PRELIMINARIES
Review of Some ODEs Results
Later on in this book, we will encounter problems where a given partial
diﬀerential is reduced to an ordinary diﬀerential function by means of a given
change of variables. Then techniques from the theory of ODE are required in
solving the transformed ODE. In this chapter, we include some of the results
from ODE theory that will be needed in our future discussions.
3 The Method of Integrating Factor
In this section, we discuss a technique for solving the ﬁrst order linear non-
homogeneous equation
y

+ p(t)y = g(t) (3.1)
where p(t) and g(t) are continuous on the open interval a < t < b.
Since p(t) is continuous, it has an antiderivative namely
_
p(t)dt. Let µ(t) =
e

p(t)dt
. Multiply Equation (3.1) by µ(t) and notice that the left hand side of
the resulting equation is the derivative of a product. Indeed,
d
dt
(µ(t)y) = µ(t)g(t).
Integrate both sides of the last equation with respect to t to obtain
µ(t)y =
_
µ(t)g(t)dt + C
Hence,
y(t) =
1
µ(t)
_
µ(t)g(t)dt +
C
µ(t)
or
y(t) = e

p(t)dt
_
e

p(t)dt
g(t)dt + Ce

p(t)dt
23
24 REVIEW OF SOME ODES RESULTS
Notice that the second term of the previous expression is just the general
solution for the homogeneous equation
y

+ p(t)y = 0
whereas the ﬁrst term is a solution to the nonhomogeneous equation. That
is, the general solution to Equation (3.1) is the sum of a particular solution of
the nonhomogeneous equation and the general solution of the homogeneous
equation.
Example 3.1
Solve the initial value problem
y

y
t
= 4t, y(1) = 5.
Solution.
We have p(t) = −
1
t
so that µ(t) =
1
t
. Multiplying the given equation by the
integrating factor and using the product rule we notice that
_
1
t
y
_

= 4.
Integrating with respect to t and then solving for y we ﬁnd that the general
solution is given by
y(t) = t
_
4dt + Ct = 4t
2
+ Ct.
Since y(1) = 5, we ﬁnd C = 1 and hence the unique solution to the IVP is
y(t) = 4t
2
+ t, 0 < t < ∞
Example 3.2
Find the general solution to the equation
y

+
2
t
y = ln t, t > 0.
Solution.
The integrating factor is µ(t) = e

2
t
dt
= t
2
. Multiplying the given equation
by t
2
to obtain
(t
2
y)

= t
2
ln t.
3 THE METHOD OF INTEGRATING FACTOR 25
Integrating with respect to t we ﬁnd
t
2
y =
_
t
2
ln tdt + C.
The integral on the right-hand side is evaluated using integration by parts
with u = ln t, dv = t
2
dt, du =
dt
t
, v =
t
3
3
obtaining
t
2
y =
t
3
3
ln t −
t
3
9
+ C
Thus,
y =
t
3
ln t −
t
9
+
C
t
2
26 REVIEW OF SOME ODES RESULTS
Practice Problems
Exercise 3.1
Solve the IVP: y

+ 2ty = t, y(0) = 0.
Exercise 3.2
Find the general solution: y

+ 3y = t + e
−3t
.
Exercise 3.3
Find the general solution: y

+
1
t
y = 3 cos t, t > 0.
Exercise 3.4
Find the general solution: y

+ 2y = cos (3t).
Exercise 3.5
Find the general solution: y

+ (cos t)y = −3 cos t.
Exercise 3.6
Given that the solution to the IVP ty

+ 4y = αt
2
, y(1) = −
1
3
exists on the
interval −∞< t < ∞. What is the value of the constant α?
Exercise 3.7
Suppose that y(t) = Ce
−2t
+ t + 1 is the general solution to the equation
y

+ p(t)y = g(t). Determine the functions p(t) and g(t).
Exercise 3.8
Suppose that y(t) = −2e
−t
+ e
t
+ sin t is the unique solution to the IVP
y

+ y = g(t), y(0) = y
0
. Determine the constant y
0
and the function g(t).
Exercise 3.9
Find the value (if any) of the unique solution to the IVP y

+ (1 + cos t)y =
1 + cos t, y(0) = 3 in the long run?
Exercise 3.10
Solve
au
x
+ bu
y
+ cu = 0
by using the change of variables s = ax + by and t = bx −ay.
3 THE METHOD OF INTEGRATING FACTOR 27
Sample Exam Questions
Exercise 3.11
Solve the initial value problem ty

= y + t, y(1) = 7.
Exercise 3.12
Show that if a and λ are positive constants, and b is any real number, then
every solution of the equation
y

+ ay = be
−λt
has the property that y →0 as t →∞. Hint: Consider the cases a = λ and
a = λ separately.
Exercise 3.13
Solve the initial-value problem y

+y = e
t
y
2
, y(0) = 1 using the substitution
u(t) =
1
y(t)
Exercise 3.14
Solve the initial-value problem ty

+ 2y = t
2
−t + 1, y(1) =
1
2
Exercise 3.15
Solve y

1
t
y = sin t, y(1) = 3. Express your answer in terms of the sine
integral, Si(t) =
_
t
0
sin s
s
ds.
28 REVIEW OF SOME ODES RESULTS
4 The Method of Separation of Variables for
ODEs
The method of separation of variables that you have seen in the theory of
ordinary diﬀerential equations has an analogue in the theory of partial dif-
ferential equations (Section 17). In this section, we review the method for
ordinary diﬀerentiable equations.
A ﬁrst order diﬀerential equation is separable if it can be written with one
variable only on the left and the other variable only on the right:
f(y)y

= g(t)
To solve this equation, we proceed as follows. Let F(t) be an antiderivative
of f(t) and G(t) be an antiderivative of g(t). Then by the Chain Rule
d
dt
F(y) =
dF
dy
dy
dt
= f(y)y

Thus,
f(y)y

−g(t) =
d
dt
F(y) −
d
dt
G(t) =
d
dt
[F(y) −G(t)] = 0
It follows that
F(y) −G(t) = C
which is equivalent to
_
f(y)y

dt =
_
g(t)dt + C
As you can see, the result is generally an implicit equation involving a func-
tion of y and a function of t. It may or may not be possible to solve this to
get y explicitly as a function of t. For an initial value problem, substitute the
values of t and y by t
0
and y
0
to get the value of C.
Remark 4.2
If F is a diﬀerentiable function of y and y is a diﬀerentiable function of t and
both F and y are given then the chain rule allows us to ﬁnd
dF
dt
given by
dF
dt
=
dF
dy
·
dy
dt
For separable equations, we are given f(y)y

=
dF
dt
and we are asked to ﬁnd
F(y). This process is referred to as “reversing the chain rule.”
4 THE METHOD OF SEPARATION OF VARIABLES FOR ODES 29
Example 4.1
Solve the initial value problem y

= 6ty
2
, y(1) =
1
25
.
Solution.
Separating the variables and integrating both sides we obtain
_
y

y
2
dt =
_
6tdt
or

_
d
dt
_
1
y
_
dt =
_
6tdt
Thus,

1
y(t)
= 3t
2
+ C
Since y(1) =
1
25
, we ﬁnd C = −28. The unique solution to the IVP is then
given explicitly by
y(t) =
1
28 −3t
2
Example 4.2
Solve the IVP yy

= 4 sin (2t), y(0) = 1.
Solution.
This is a separable diﬀerential equation. Integrating both sides we ﬁnd
_
d
dt
_
y
2
2
_
dt = 4
_
sin (2t)dt
Thus,
y
2
= −4 cos (2t) + C
Since y(0) = 1, we ﬁnd C = 5. Now, solving explicitly for y(t) we ﬁnd
y(t) = ±

−4 cos t + 5
Since y(0) = 1, we have y(t) =

−4 cos t + 5. The interval of existence of
the solution is the interval −∞< t < ∞
30 REVIEW OF SOME ODES RESULTS
Practice Problems
Exercise 4.1
Solve the (separable) diﬀerential equation
y

= te
t
2
−ln y
2
.
Exercise 4.2
Solve the (separable) diﬀerential equation
y

=
t
2
y −4y
t + 2
.
Exercise 4.3
Solve the (separable) diﬀerential equation
ty

= 2(y −4).
Exercise 4.4
Solve the (separable) diﬀerential equation
y

= 2y(2 −y).
Exercise 4.5
Solve the IVP
y

=
4 sin (2t)
y
, y(0) = 1.
Exercise 4.6
Solve the IVP:
yy

= sin t, y(
π
2
) = −2.
Exercise 4.7
Solve the IVP:
y

+
1
y + 1
= 0, y(1) = 0.
Exercise 4.8
Solve the IVP:
y

−ty
3
= 0, y(0) = 2.
4 THE METHOD OF SEPARATION OF VARIABLES FOR ODES 31
Exercise 4.9
Solve the IVP:
y

= 1 + y
2
, y(
π
4
) = −1.
Exercise 4.10
Solve the IVP:
y

= t −ty
2
, y(0) =
1
2
.
32 REVIEW OF SOME ODES RESULTS
Sample Exam Questions
Exercise 4.11
For what values of the constants α, y
0
, and integer n is the function y(t) =
(4 + t)

1
2
a solution of the initial value problem?
y

+ αy
n
= 0, y(0) = y
0
.
Exercise 4.12
Solve the equation 3u
y
+ u
xy
= 0 by using the substitution v = u
y
.
Exercise 4.13
Solve the IVP
(2y −sin y)y

= sin t −t, y(0) = 0.
Exercise 4.14
State an initial value problem, with initial condition imposed at t
0
= 2,
having implicit solution y
3
+ t
2
+ sin y = 4.
Exercise 4.15
Can the diﬀerential equation
dy
dx
= x
2
−xy
be solved by the method of separation of variables? Explain.
5 SECOND ORDER LINEAR ODES 33
5 Second Order Linear ODEs
When solving second order partial diﬀerential equations such as the heat,
wave, and Laplace’s equations using the method of separation of variables
for PDEs one ends up confronting second order linear ODEs. Thus, it is
deemed necessary to review some of the techniques usided in solving second
order linear ordinary diﬀerential equations which we do in this section.
We start ﬁrst by considering the second order linear ODE with constant
coeﬃcients given by
ay

+ by

+ cy = 0 (5.1)
where a, b and c are constants with a = 0.
Notice ﬁrst that for b = 0 and c = 0 the function y

is a constant multiple
of y. So it makes sense to look for a function with such property. One such
function is y(t) = e
rt
. Substituting this function into (5.1) leads to
ay

+ by

+ cy = ar
2
e
rt
+ bre
rt
+ ce
rt
= (ar
2
+ br + c)e
rt
= 0
Since e
rt
> 0 for all t, the previous equation leads to
ar
2
+ br + c = 0 (5.2)
Thus, a function y(t) = e
rt
is a solution to (5.1) when r satisﬁes equation
(5.2). We call (5.2) the characteristic equation for (5.1) and the polyno-
mial C(r) = ar
2
+ br + c is called the characteristic polynomial.
The characteristic equation is a quadratic equation. Thus, this equation can
have two distinct real solutions, two equal solutions, or two conjugate com-
plex solutions depending on the sign of the discriminant b
2
−4ac. Hence, we
consider the following three cases:
Case 1: b
2
−4ac > 0.
In this case, equation (5.2) have two distinct real roots r
1
=
−b−

b
2
−4ac
4a
and
r
2
=
−b+

b
2
−4ac
4a
. The general solution to (5.1) is given by
y(t) = c
1
e
r
1
t
+ c
2
e
r
2
t
where c
1
and c
2
are arbitrary constants.
Example 5.1
Solve the initial value problem
y

−y

−6y = 0, y(0) = 1, y

(0) = 2.
Describe the behavior of the solution y(t) as t →−∞ and t →∞.
34 REVIEW OF SOME ODES RESULTS
Solution.
The characteristic polynomial is C(r) = r
2
−r−6 = (r−3)(r+2) so that the
characteristic equation r
2
−r −6 = 0 has the solutions r
1
= 3 and r
2
= −2.
The general solution is then given by
y(t) = c
1
e
3t
+ c
2
e
−2t
.
Taking the derivative to obtain
y

(t) = 3c
1
e
3t
−2c
2
e
−2t
.
The conditions y(0) = 1 and y

(0) = 2 lead to the system
c
1
+ c
2
= 1
3c
1
−2c
2
= 2.
Solving this system by the method of elimination we ﬁnd c
1
=
4
5
and c
2
=
1
5
.
Hence, the unique solution to the initial value problem is
y(t) =
1
5
(4e
3t
+ e
−2t
).
As t → −∞, e
3t
→ 0 and e
−2t
→ ∞. Thus, y(t) → ∞. Similarly, y(t) → ∞
as t →∞
Case 2: b
2
−4ac = 0.
In this case, the characteristic equation has the single root r = −
b
2a
. The
general solution to (5.1) is given by
y(t) = c
1
e

b
2a
t
+ c
2
te

b
2a
t
where c
1
and c
2
are arbitrary constants.
Example 5.2
Solve the initial value problem: y

+ 2y

+ y = 0, y(0) = 1, y

1
(0) = −1.
Solution.
The characteristic equation r
2
+2r+1 = 0 has a repeated root: r
1
= r
2
= −1.
Thus, the general solution is given by
y(t) = c
1
e
−t
+ c
2
te
−t
.
5 SECOND ORDER LINEAR ODES 35
The two conditions y(0) = 1 and y

(0) = −1 lead to c
2
= 1 and c
1
= 0.
Hence, the unique solution is y(t) = e
−t
Case 3: b
2
−4ac < 0.
In this case, the complex roots of equation (5.1) are given by
r
1,2
=
−b ±i

4ac −b
2
2a
where i =

−1. The general solution is given by
y(t) = e
αt
(c
1
cos βt + c
2
sin βt)
where α = −
b
2a
, β =

4ac−b
2
2a
, and c
1
and c
2
are real numbers.
Example 5.3
Solve the initial value problem
y

−10y

+ 29y = 0, y(0) = 1, y

(0) = 3.
Solution.
The characteristic equation r
2
− 10r + 29 = 0 has the complex roots r
1,2
=
5 ±2i. Thus, the general solution is given by the expression
y(t) = e
5t
(c
1
cos 2t + c
2
sin 2t).
Finding y

we obtain
y

(t) = e
5t
[(5c
1
+ 2c
2
) cos 2t + (5c
2
−2c
1
) sin 2t].
The initial conditions yield c
1
= 1 and c
2
= −1. Thus, the unique solution
to the initial value problem is
y(t) = e
5t
(cos 2t −sin 2t)
The Eigenvalue Problem
Consider the question of ﬁnding a twice diﬀerentaible function u satisfying
the ordinary diﬀerential equation
d
2
u
dx
2
= λu, a < x < b.
36 REVIEW OF SOME ODES RESULTS
subject to the boundary conditions u(a) = u(b) = 0. This problem is referred
to as the eigenvalue problem for the following reason: Deﬁne the function
L ≡
d
2
dx
2
. Then the given equation can be written as Lu = λu. In linear
algebra, λ is called an eigenvalue of L with corresponding eigenvector u.
Diﬀerent solutions to the eigenvalue problem are obtained depending on the
sign of λ. Suppose ﬁrst that λ = 0. Then u(x) = ax+b for arbitrary constants
a and b. Using the boundary conditions we ﬁnd a = b = 0. Hence, u ≡ 0.
Suppose that λ > 0. Then u(x) = Ae

λx
+ Be

λx
. Again, the boundary
conditions imply that u ≡ 0.
Now, suppose that λ < 0. Then u(x) = Acos

−λx + Bsin

−λx. Using
the condition u(0) = 0 to obtain A = 0. Using the condition u(1) = 0
and assuming we are looking for non-trivial solution u we expect to have
sin

−λ = 0. This happens when λ = λ
n
= −(nπ)
2
where n ∈ N. We call λ
n
an eigenvalue with corresponding eigenfunction u
n
(x) = A
n
sin nπx.
Finally, using the principle of superposition we ﬁnd that the general solution
to the eigenvalue problem is given by
u(x) =

n=1
A
n
u
n
(x)
where the convergence is pointwise convergence (See Section 2).
Euler Equations
A second order linear diﬀerential equations of the form
ax
2
y

+ bxy

+ cy = 0
where a, b, c are constants is called an Euler equation.
To solve Euler equation, one starts with solutions of the form y = x
r
(with
x > 0) where r is to be determined. Plugging this into the diﬀerential
equation to get
ax
2
r(r −1)x
r−2
+ bxrx
r−1
+ cx
r
=0
(ar
2
−ar + br + c)x
r
=0
ar
2
−(a −b)r + c =0
This last equation is a quadratic equation in r and so we will have three cases
to look at : Real distinct roots, double roots, and complex conjugate roots.
5 SECOND ORDER LINEAR ODES 37
If the quadratic equation has two distinct real roots r
1
and r
2
then the general
solution is given by
y(x) = Ax
r
1
+ Bx
r
2
.
If the quadratic equation has two equal roots r
1
= r
2
= r then the general
solution is given by
y(x) = x
r
(A + Bln x).
If the quadratic equation has two complex conjugate solutions r
1,2
= α ±iβ
then the general solution is given by
y(x) = x
α
(Acos (β ln x) + Bsin (β ln x)).
Example 5.4
Solve the initial value problem
2x
2
y

+ 3xy

−15y = 0
y(1) = 0, y

(1) = 1.
Solution.
Letting y = x
r
we obtain the quadratic equation 2r
2
+ r − 15 = 0 whose
roots are r
1
=
5
2
and r
2
= −3. Hence, the general solution is given by
y(x) = Ax
5
2
+ Bx
−3
.
The condition y(1) = 0 implies A + B = 0. The condition y

(1) = 1 implies
5
2
A − 3B = 1. Solving this system of two unknowns we ﬁnd A =
2
11
and
B = −
2
11
. Hence, the unique solution is given by
y =
2
11
x
5
2

2
11
x
−3
Second Order Linear Inhomogeneous ODE: The Method of Unde-
termined Coeﬃcients
We consider the nonhomogeneous second order
ay

+ by

+ cy = g(t), a < t < b.
We know that the general solution has the structure
y(t) = c
1
y
1
(t) + c
2
y
2
(t) + y
p
(t)
38 REVIEW OF SOME ODES RESULTS
where y
p
(t) is a particular solution to the nonhomogeneous equation. We
will write y(t) = y
h
(t) + y
p
(t) where y
h
(t) = c
1
y
1
(t) + c
2
y
2
(t).
One way to ﬁnding y
p
is by using the methos of undetermined coeﬃcients.
The idea behind the method of undetermined coeﬃcients is to look for y
p
(t)
which is of a form like that of g(t). This is possible only for special functions
g(t), but these special cases arise quite frequently in applications.
We will assume that g(t) being simple means it is some combination of terms
like e
rt
, cos (kt), sin (kt), and polynomials a
n
t
n
+a
n−1
t
n−1
+· · · a
1
t +a
0
. (Note
that if both cosine and sine terms are present, if they have the same argument
kt they can be treated as one. But if they have diﬀerent arguments they must
be treated separately, each resulting in a combination of sine and cosine terms
in y
p
.) Based on those terms we will put together a candidate y
p
that has some
constants in it we need to solve for: Those are the undetermined coeﬃcients
this method is named for.
In the following table we list examples of g(t) along with the corresponding
form of the particular solution.
Form of g(t) Form of y
p
(t)
a
n
t
n
+ a
n−1
t
n−1
+· · · + a
1
t + a
0
t
r
[A
n
t
n
+ A
n−1
t
n−1
+· · · + A
1
t + A
0
[a
n
t
n
+ a
n−1
t
n−1
+· · · + a
1
t + a
0
]e
αt
t
r
[A
n
t
n
+ A
n−1
t
n−1
+· · · + A
1
t + A
0
]e
αt
[a
n
t
n
+ a
n−1
t
n−1
+· · · + a
1
t + a
0
] cos αt t
r
[(A
n
t
n
+ A
n−1
t
n−1
+· · · + A
1
t + A
0
) cos αt
or +(B
n
t
n
+ B
n−1
t
n−1
+· · · + B
1
t + B
0
) sin αt]
[a
n
t
n
+ a
n−1
t
n−1
+· · · + a
1
t + a
0
] sin αt
e
αt
[a
n
t
n
+ a
n−1
t
n−1
+· · · + a
1
t + a
0
] sin βt t
r
[(A
n
t
n
+ A
n−1
t
n−1
+· · · + A
1
t + A
0
)e
αt
cos βt
or +(B
n
t
n
+ B
n−1
t
n−1
+· · · + B
1
t + B
0
)e
αt
sin βt]
e
αt
[a
n
t
n
+ a
n−1
t
n−1
+· · · + a
1
t + a
0
] cos βt
The number r is chosen to be the smallest nonnegative integer such that
no term in the assumed form is a solution of the homogeneous equation
ay

+ by

+ cy = 0. The value of r will be 0, 1, or 2.
Example 5.5
List an appropriate form for a particular solution of
(a) y

+ 4y = t
2
e
3t
.
(b) y

+ 4y = te
2t
cos t.
(c) y

+ 4y = 2t
2
+ 5 sin 2t + e
3t
.
(d) y

+ 4y = t
2
cos 2t.
5 SECOND ORDER LINEAR ODES 39
Solution.
The general solution to the homogeneous equation is y
h
(t) = c
1
cos t+c
2
sin t.
(a) For g(t) = t
2
e
3t
, an appropriate particular solution has the form y
p
(t) =
t
r
(A
2
t
2
+A
1
t +A
0
)e
3t
. We take r = 0 since no term in the assumed form for
y
p
is present in the expression of y
h
(t). Thus
y
p
(t) = (A
2
t
2
+ A
1
t + A
0
)e
3t
(b) An appropriate form is
y
p
(t) = t
r
[(A
1
t + A
0
)e
2t
cos t + (B
1
t + B
0
)e
2t
sin t]
We take r = 0 since no term in the assumed form for y
p
is present in the
expression of y
h
(t). Thus
y
p
(t) = (A
1
t + A
0
)e
2t
cos t + (B
1
t + B
0
)e
2t
sin t
(c)
y
p
(t) = A
2
t
2
+ A
1
t + A
0
+ B
0
t cos 2t + C
0
t sin 2t + D
0
e
3t
(d)
y
p
(t) = t(A
2
t
2
+ A
1
t + A
0
) cos 2t + t(B
2
t
2
+ B
1
t + B
0
) cos 2t
Example 5.6
Find the general solution of
y

−2y

−3y = 4t −5 + 6te
2t
Solution.
The characteristic equation of the homogeneous equation is r
2
−2r −3 = 0
with roots r
1
= −1 and r
2
= 3. Thus,
y
h
(t) = c
1
e
−t
+ c
2
te
3t
A guess for the particular solution is y
p
(t) = At +B+Cte
2t
+De
2t
. Inserting
this into the diﬀerential equation leads to
−3At −2A −3B −3Cte
2t
+ (2C −D)e
2t
= 4t −5 + 6te
2t
From this identity we obtain −3A = 4 so that A = −
4
3
. Also, −2A−3B = −5
so that B =
23
9
. Since −3C = 6 we ﬁnd C = −2. From 2C −3D = 0 we ﬁnd
D = −
4
3
. It follows that
y(t) = c
1
e
−t
+ c
2
te
3t

4
3
t +
23
9

_
2t +
4
3
_
e
2t
40 REVIEW OF SOME ODES RESULTS
Practice Problems
Exercise 5.1
Solve the initial value problem
y

−4y

+ 3y = 0, y(0) = −1, y

(0) = 1
Describe the behavior of the solution y(t) as t →−∞ and t →∞.
Exercise 5.2
Solve the initial value exercise
y

+ 4y

+ 2y = 0, y(0) = 0, y

(0) = 4
Describe the behavior of the solution y(t) as t →−∞ and t →∞.
Exercise 5.3
Solve the initial value problem
2y

−y = 0, y(0) = −2, y

(0) =

2
Describe the behavior of the solution y(t) as t →−∞ and t →∞.
Exercise 5.4
Find a homogeneous second-order linear ordinary diﬀerential equation whose
general solution is y(t) = c
1
e
2t
+ c
2
e
−t
.
Exercise 5.5
Solve the IVP
9y

−6y

+ y = 0, y(3) = −2, y

(3) = −
5
3
Exercise 5.6
Solve the IVP
25y

+ 20y

+ 4y = 0, y(5) = 4e
−2
, y

(5) = −
3
5
e
−2
Exercise 5.7
The graph of a solution y(t) of the diﬀerential equation 4y

+ 4y

+ y = 0
passes through the points (1, e

1
2
) and (2, 0). Determine y(0) and y

(0).
5 SECOND ORDER LINEAR ODES 41
Exercise 5.8
Find the general solution of y

−6y

+ 9y = 0.
Exercise 5.9
Solve the IVP
y

+ 2y

+ 2y = 0, y(0) = 3, y

(0) = −1
Exercise 5.10
Solve the IVP
2y

−2y

+ y = 0, y(−π) = 1, y

(−π) = −1
Exercise 5.11
Find the general solution of
y

−y

+ y = 2 sin 3t
Exercise 5.12
Find the general solution of
y

+ 4y

−2y = 2t
2
−3t + 6
42 REVIEW OF SOME ODES RESULTS
Sample Exam Questions
Exercise 5.13
Find the general solution to the following diﬀerential equation.
x
2
y

−7xy

+ 16y = 0.
Exercise 5.14
Find the general solution to the following diﬀerential equation.
x
2
y

+ 3xy

+ 4y = 0.
Exercise 5.15
Consider the diﬀerential equation
d
2
y
dx
2
+ λy = 0.
Determine the eigenvalues λ and the corresponding eigenfunctions if y satis-
ﬁes the following boundary conditions:
(a) y(0) = y(π) = 0
(b) y(0) = y

(L) = 0
(c) y

(0) = y(1) = 0.
Exercise 5.16
Show by direct computation that the eigenvalue problems
(ky

(x))

+ λy(x) = 0
with the following boundary conditions have no negative eigenvalues λ :
(a) y(0) = y(L) = 0
(b) y

(0) = y

(L) = 0
(c) y(L) = y(−L), y

(L) = y

(−L) and k(−L) = k(L).
Exercise 5.17
Solve the initial-value problem: 2y

+ 5y

−3y = 0, y(0) = 2, y

(0) = 1.
Exercise 5.18
Find the general solution of
y

−y

= 5e
t
−sin 2t
Exercise 5.19
Solve using undetermined coeﬃcients:
y

+ y

−2y = t + sin 2t, y(0) = 1, y

(0) = 0
Introduction to PDEs
Many ﬁelds in engineering and the physical sciences require the study of ODE
and PDE. Examples of those ﬁelds are acoustics, aerodynamics, elasticity,
electrodynamics, ﬂuid dynamics, geophysics (seismic wave propagation), heat
transfer, meteorology, oceanography, optics, petroleum engineering, plasma
physics (ionized liquids and gases), quantum mechanics.
So the study of partial diﬀerential equation is of great importance to the
above mentioned ﬁelds. The purpose of this chapter is to introduce the
reader to the basic terms of partial diﬀerential equations.
6 The Basic Concepts
The goal of this section is to introduce the reader to the basic concepts and
notations that will be used in the remainder of this book.
A diﬀerential equation is an equation that involves an unknown scalar
function (the dependent variable) and one or more of its derivatives. For
example,
d
2
y
dx
2
−5
dy
dx
+ 3y = −3 (6.1)
or
∂u
∂t

2
u
∂x
2

2
u
∂y
2
+ u = 0. (6.2)
If the unknown function is a function in one single variable then the diﬀer-
ential equation is called an ordinary diﬀerential equation. An example
of an ordinary diﬀerential equation is Equation (6.1). In contrast, when the
unknown function is a function of two or more independent variables then
the diﬀerential equation is called a partial diﬀerential equation, in short
PDE. Equation (6.2) is an example of a partial diﬀerential equation. In this
book we will be focusing on partial diﬀerential equations.
43
44 INTRODUCTION TO PDES
Example 6.1
Identify which variables are dependent variable or independent variable(s)
for the following diﬀerential equations.
(a)
d
4
y
dx
4
−x
2
+ y = 0
(b) u
tt
+ xu
tx
= 0.
(c) x
dx
dt
= 4.
(d)
∂y
∂u
−4
∂y
∂v
= u + 3y.
Solution.
(a) Independent variable is x and the dependent variable is y.
(b) Independent variables are x and t and the dependent variable is u.
(c) Independent variable is t and the dependent variable is x.
(d) Independent variable are u and v and the dependent variable is y
Example 6.2
Classify the following as either ODE or PDE.
(a) u
t
= c
2
u
xx
.
(b) y

−4y

+ 5y = 0.
(c) u
t
+ cu
x
= 5.
Solution.
(a) PDE (b) ODE (c) PDE
The order of a partial diﬀerential equation is the highest order derivative
occurring in the equation. Thus, (6.2) is a second order partial diﬀerential
equation.
Example 6.3
Find the order of each of the following partial diﬀerential equations:
(a) xu
x
+ yu
y
= x
2
+ y
2
(b) uu
x
+ u
y
= 2
(c) u
tt
−c
2
u
xx
= f(x, t)
(d) u
t
+ uu
x
+ u
xxx
= 0
(e) u
tt
+ u
xxxx
= 0.
Solution.
(a) First order (b) First order (c) Second order (d) Third order (e) Fourth
order
6 THE BASIC CONCEPTS 45
A partial diﬀerential equation is called linear if it is linear in the unknown
function and all its derivatives with coeﬃcients depend only on the indepen-
dent variables. For example, a ﬁrst order linear partial diﬀerential equation
has the form
A(x, y)u
x
+ B(x, y)u
y
+ C(x, y)u = D(x, y)
whereas a second order linear partial diﬀerential equation has the form
A(x, y)u
xx
+B(x, y)u
xy
+C(x, y)u
yy
+D(x, y)u
x
+E(x, y)u
y
+F(x, y)u = G(x, y).
A partial diﬀerential equation is called quasi-linear if it is linear in the
highest-order derivatives which appear in the equation (regardless of the
manner in which lower-order derivatives and unknown functions occur in the
equation). For example, a ﬁrst order quasi-linear partial diﬀerential equation
has the form
A(x, y, u)u
x
+ B(x, y, u)u
y
= C(x, y, u)
whereas a second order quasi-linear partial diﬀerential equation has the form
A(x, y, u, u
x
, u
y
)u
xx
+B(x, y, u, u
x
, u
y
)u
xy
+C(x, y, u, u
x
, u
y
)u
yy
= D(x, y, u, u
x
, u
y
).
A partial diﬀerential equation is semi-linear if it is quasi-linear and the
coeﬃcients of the highest-order derivatives are functions of independent vari-
ables only. For example, a ﬁrst order semi-linear partial diﬀerential equation
has the form
A(x, y)u
x
+ B(x, y)u
y
= C(x, y, u)
whereas a second order semi-linear partial diﬀerential equation has the form
A(x, y)u
xx
+ B(x, y)u
xy
+ C(x, y)u
yy
= D(x, y, u, u
x
, u
y
).
Note that linear and semi-linear partial diﬀerential equations are special cases
of quasi-linear equations.
A partial diﬀerential equation that is not linear is called nonlinear. For
example, u
2
x
+ 2u
xy
= 0.
As for ODEs, linear PDEs are usually simpler to analyze/solve than nonlinear
PDEs.
46 INTRODUCTION TO PDES
Example 6.4
Determine whether the given PDE is linear, quasilinear, semilinear, or non-
linear:
(a) xu
x
+ yu
y
= x
2
+ y
2
(b) uu
x
+ u
y
= 2
(c) u
tt
−c
2
u
xx
= f(x, t)
(d) u
t
+ uu
x
+ u
xxx
= 0
(e) u
2
tt
+ u
xxxx
= 0.
Solution.
(a) Linear, quasilinear, semilinear.
(b) Quasilinear, nonlinear.
(c) Linear, quasilinear, semilinear.
(d) Quasilinear, semilinear, nonlinear.
(e) Quasilinear, semilinear, nonlinear
A more precise deﬁnition of a linear diﬀerential equation begins with the
concept of a linear diﬀerential operator L. The operator L is assembled
by summing the basic partial derivative operators, with coeﬃcients depend-
ing on the independent variables. The operator acts on suﬃciently smooth
functions depending on the relevant independent variables. Linearity im-
poses two key requirements:
L[u + v] = L[u] + L[v] and L[αu] = αL[u],
for any two (suﬃciently smooth) functions u, v and any constant α.
Example 6.5
Deﬁne a linear diﬀerential operator for the PDE
u
t
= c
2
u
xx
.
Solution.
Let L[u] = u
t
−c
2
u
xx
. Then one can easily check that L[u +v] = L[u] +L[v]
and L[αu] = αL[u]
A linear partial diﬀerential equation is called homogeneous if every term
of the equation involves the unknown function or its partial derivatives. A
linear partial diﬀerential equation that is not homogeneous is called nonho-
mogeneous. In this case, there is a term in the equation that involves only
6 THE BASIC CONCEPTS 47
the independent variables.
A homogeneous linear partial diﬀerential equation has the form
L[u] = 0
where L is a linear diﬀerential operator.
Example 6.6
Determine whether the equation is homogeneous or nonhomogeneous:
(a) xu
x
+ yu
y
= x
2
+ y
2
.
(b) u
tt
= c
2
u
xx
.
(c) u
xx
+ u
yy
= 0.
Solution.
(a) Nonhomogeneous because of x
2
+ y
2
.
(b) Homogeneous.
(c) Homogeneous
Finally, we shall be employing a few basic notational conventions regard-
ing the variables that appear in our diﬀerential equations. We always use
t to denote time, while x, y, z will represent (Cartesian) space coordinates.
Polar coordinates r, θ will also be used when needed, and our notational con-
ventions appear at the appropriate places in the exposition.
An equilibrium equation models an unchanging physical system, and so
only involves the space variables. The time variable t appears when mod-
eling dynamical , meaning time-varying, processes. Both time and space
coordinates are independent variables.
48 INTRODUCTION TO PDES
Practice Problems
Exercise 6.1
Classify the following equations as either ODE or PDE.
(a) (y

)
4
+
t
2
(y

)
2
+4
= 0
(b)
∂u
∂x
+ y
∂u
∂y
=
y−x
y+x
(c) y

−4y = 0
Exercise 6.2
Write the equation
u
xx
+ 2u
xy
+ u
yy
= 0
in the coordinates s = x, t = x −y.
Exercise 6.3
Write the equation
u
xx
−2u
xy
+ 5u
yy
= 0
in the coordinates s = x + y, t = 2x.
Exercise 6.4
For each of the following PDEs, state its order and whether it is linear or
nonlinear. If it is linear, also state whether it is homogeneous or inhomoge-
neous:
(a) uu
x
+ x
2
u
yyy
+ sin x = 0
(b) u
x
+ e
x
2
u
y
= 0
(c) u
tt
+ (sin y)u
yy
−e
t
cos y = 0.
Exercise 6.5
For each of the following PDEs, determine its order and whether it is linear
or not. For linear PDEs, state also whether the equation is homogeneous or
not. For nonlinear PDEs, circle all term(s) that are not linear.
(a) x
2
u
xx
+ e
x
u = xu
xyy
(b) e
y
u
xxx
+ e
x
u = −sin y + 10xu
y
(c) y
2
u
xx
+ e
x
uu
x
= 2xu
y
+ u
(d) u
x
u
xxy
+ e
x
uu
y
= 5x
2
u
x
(e) u
t
= k
2
(u
xx
+ u
yy
) + f(x, y, t).
6 THE BASIC CONCEPTS 49
Exercise 6.6
Which of the following PDEs are linear?
(a) Laplace’s equation: u
xx
+ u
yy
= 0.
(b) Convection (transport) equation: u
t
+ cu
x
= 0.
(c) Minimal surface equation: (1+Z
2
y
)Z
xx
−2Z
x
Z
y
Z
xy
+(1+Z
2
x
)Z
yy
= 0.
(d) Korteweg-Vries equation: u
t
+ 6uu
x
= u
xxx
.
Exercise 6.7
Classify the following diﬀerential equations as ODEs or PDEs, linear or
nonlinear, and determine their order. For the linear equations, determine
whether or not they are homogeneous.
(a) The diﬀusion equation for u(x, t) :
u
t
= ku
xx
.
(b) The wave equation for w(x, t) :
w
tt
= c
2
w
xx
.
(c) The thin ﬁlm equation for h(x, t) :
h
t
= −(hh
xxx
)
x
.
(d) The forced harmonic oscillator for y(t) :
y
tt
+ ω
2
y = F cos (ωt).
(e) The Poisson Equation for the electric potential Φ(x, y, z) :
Φ
xx
+ Φ
yy
+ Φ
zz
= 4πρ(x, y, z).
where ρ(x, y, z) is a known charge density.
(f) Burger’s equation for h(x, t) :
h
t
+ hh
x
= νh
xx
.
Exercise 6.8
Write down the general form of a linear ﬁrst order diﬀerential equation of a
function in three variables.
50 INTRODUCTION TO PDES
Exercise 6.9
Give the orders of the following PDEs, and classify them as linear or non-
linear. If the PDE is linear, specify whether it is homogeneous or inhomoge-
neous.
(a) x
2
u
xxy
+ y
2
u
yy
−log (1 + y
2
)u = 0
(b) u
x
+ u
3
= 1
(c) u
xxyy
+ e
x
u
x
= y
(d) uu
xx
+ u
yy
−u = 0
(e) u
xx
+ u
t
= 3u.
Exercise 6.10
Consider the second-order PDE
u
xx
+ 4u
xy
+ 4u
yy
= 0.
Use the change of variables v(x, y) = y − 2x and w(x, y) = x to show that
u
ww
= 0.
6 THE BASIC CONCEPTS 51
Sample Exam Questions
Exercise 6.11
Write the one dimensional wave equation u
tt
= c
2
u
xx
in the coordinates
v = x + ct and w = x −ct.
Exercise 6.12
Write the PDE
u
xx
+ 2u
xy
−3u
yy
= 0
in the coordinates v(x, y) = y −3x and w(x, y) = x + y.
Exercise 6.13
Write the PDE
au
x
+ bu
y
= 0
in the coordinates s(x, y) = ax+by and t(x, y) = bx−ay. Assume a
2
+b
2
> 0.
Exercise 6.14
Write the PDE
u
x
+ u
y
= 1
in the coordinates s = x + y and t = x −y.
Exercise 6.15
Write the PDE
au
t
+ bu
x
= u, a, b = 0
in the coordinates v = ax −bt and w =
1
a
t.
52 INTRODUCTION TO PDES
7 Solutions and Related Topics
By a classical solution or strong solution to a partial diﬀerential equation
we mean a function that satisﬁes the equation. To solve a PDE is to ﬁnd all
its classical solutions. In the case of only two independent variables x and y,
a solution u(x, y) is visualized geometrically as a surface, called a solution
surface or an integral surface in the (x, y, u) space.
A formula that expresses all the solutions of a PDE is called the general
solution of the equation.
Example 7.1
Show that u(x, t) = e
−λ
2
α
2
t
(cos λx − sin λx) is a solution to the equation
u
t
−α
2
u
xx
= 0.
Solution.
Since
u
t
−α
2
u
xx
= −λ
2
α
2
e
−λ
2
α
2
t
(cos λx −sin λx) −α
2
(−λ
2
cos λx +λ
2
sin λx) = 0
the given function is a solution to the given equation
Example 7.2
Find the general solution of u
xy
= 0.
Solution.
Integrating ﬁrst we respect to y we ﬁnd u
x
(x, y) = f(x), where f is an
arbitrary diﬀerentiable function. Integrating u
x
with respect to x we ﬁnd
u(x, y) = f(x) + g(y), where g is an arbitrary diﬀerentiable function
Note that the general solution in the previous example involves two arbitrary
functions. In general, the general solution of a partial diﬀerential equation
is an expression that involves arbitrary functions. This is in contrast to the
general solution of an ordinary diﬀerential equation which involves arbitrary
constants.
Usually, a classical solution enjoys properties such as smootheness (i.e. dif-
ferentiability) and continuity. However, in the theory of nonlinear pdes, there
are solutions that do not require the smoothness property. Such solutions are
called weak solutions or generalized solutions. We illustrate this con-
cept using equations rather than pdes. Consider the equation x
2
− y
2
= 0.
7 SOLUTIONS AND RELATED TOPICS 53
The function y = x is a classical solution of this equation. This solution is
inﬁnitely diﬀerentiable function. On the other hand, the function y = |x| is
also a solution to the given equation. However, this solution is not diﬀeren-
tiable at 0. We call such a solution a weak solution. In this book, the word
solution will refer to a classical solution.
Example 7.3
Show that u(x, t) = t +
1
2
x
2
is a classical solution to the PDE
u
t
= u
xx
. (7.1)
Solution.
Assume that the domain of deﬁnition of u is D ⊂ R
2
. Since u, u
t
, u
x
, u
tx
, u
xx
exist and are continuous in D(i.e., u is smooth in D) and u satisﬁes equation
(7.1), we conclude that u is a classical solution to the given PDE
Now, consider the linear diﬀerential operator L as deﬁned in the previous
section. The deﬁning properties of linearity immediately imply the key facts
concerning homogeneous linear (diﬀerential) equations.
Theorem 7.1
The sum of two solutions to a homogeneous linear diﬀerential equation is
again a solution, as is the product of a solution by any constant.
Proof.
Let u
1
, u
2
be solutions, meaning that L[u
1
] = 0 and L[u
2
] = 0. Then, thanks
to linearity,
L[u
1
+ u
2
] = L[u
1
] + L[u
2
] = 0,
and hence their sum u
1
+u
2
is a solution. Similarly, if α is any constant, and
u any solution, then
L[αu] = αL[u] = α0 = 0,
and so the scalar multiple αu is also a solution
The following result is known as the superposition principle for homo-
geneous linear equations.
Theorem 7.2
If u
1
, · · · , u
n
are solutions to a common homogeneous linear partial diﬀeren-
tial equation L[u] = 0, then the linear combination u = c
1
u
1
+ · · · + c
n
u
n
is
a solution for any choice of constants c
1
, · · · , c
n
.
54 INTRODUCTION TO PDES
Proof.
The key fact is that, thanks to the linearity of L, for any suﬃciently smooth
functions u
1
, · · · , u
n
and any constants c
1
, · · · , c
n
,
L[u] =L[c
1
u
1
+· · · + c
n
u
n
] = L[c
1
u
1
+· · · + c
n−1
u
n−1
] + L[c
n
u
n
]
=· · · = L[c
1
u
1
] +· · · + L[c
n
u
n
] = c
1
L[u
1
] +· · · + c
n
L[u
n
].
In particular, if the functions are solutions, so L[u
1
] = 0, · · · , L[u
n
] = 0, then
the right hand side of the above equation vanishes, proving that u is also a
solution to the homogeneous equation L[u] = 0
In physical applications, homogeneous linear equations model unforced sys-
tems that are subject to their own internal constraints. External forcing
is represented by an additional term that does not involve the dependent
variable. This results in the nonhomogeneous equation
L[u] = f
where L is a linear partial diﬀerential operator, u is the dependent variable,
and f is a given non-zero function of the independent variables alone.
You already learned the basic philosophy for solving of inhomogeneous linear
equations in your study of elementary ordinary diﬀerential equations. Step
one is to determine the general solution to the homogeneous equation. Step
two is to ﬁnd a particular solution to the inhomogeneous version. The general
solution to the inhomogeneous equation is then obtained by adding the two
together. Here is the general version of this procedure:
Theorem 7.3
Let v
i
be a particular solution to the inhomogeneous linear equation L[u
i
] =
f. Then the general solution to L[u] = f is given by u = u
i
+u
h
, where u
h
is
an arbitrary solution to the corresponding homogeneous equation L[u
h
] = 0.
Proof.
Let us ﬁrst show that u = u
i
+ u
h
is also a solution whenever L[u] = 0. By
linearity,
L[u] = L[u
i
+ u
h
] = L[u
i
] + L[u
h
] = f + 0 = f.
To show that every solution to the inhomogeneous equation can be expressed
in this manner, suppose u satisﬁes L[u] = f. Set u
h
= u − u
i
. Then, by
linearity,
L[u
h
] = L[u −u
i
] = L[u] −L[u
i
] = 0,
7 SOLUTIONS AND RELATED TOPICS 55
and hence u
h
is a solution to the homogeneous diﬀerential equation. Thus,
u = u
i
+ u
h
has the required form
In physical applications, one can interpret the particular solution u
i
as a
response of the system to the external forcing function, while the solution
u
h
to the homogeneous equation represents the system’s internal, unforced
motion. The general solution to a linear inhomogeneous equation is thus a
combination of the external and internal responses.
As you have noticed by now, one solution of a linear PDE leads to the cre-
ation of lots of solutions. In contrast, nonlinear equations are much tougher
to deal with, for example, knowledge of several solutions is of scant help
constructing others. Indeed, even ﬁnding one solution to a nonlinear partial
diﬀerential equation can be quite a challenge.
In this introductory course, we will primarily − but not exclusively − con-
centrate on analyzing the most basic linear partial diﬀerential equations. But
we will have occasion to brieﬂy foray into the nonlinear realm, to appreciate
some recent developments in this fascinating area of contemporary research
and applications.
As observed above, a general solution of a partial diﬀerential equation has
inﬁnitely many solutions. In almost all cases, this general solution is of little
use since it has to satisfy other supplementary conditions, usually called ini-
tial or boundary conditions. These conditions determine the unique solution
of interest.
A boundary value problem is a partial diﬀerential equation where either
the unknown function or its derivatives have values assigned on the physical
boundary of the domain in which the problem is speciﬁed. These conditions
are called boundary conditions. For example,
u
xx
+ u
yy
=0 if 0 < x, y < 1
u(x, 0) = u(x, 1) =0 if 0 < x < 1
u
x
(0, y) = u
x
(1, y) =0 if 0 < y < 1.
There are three types of boundary conditions which arise frequently in for-
mulating physical problems:
1. Dirichlet Boundary Conditions: In this case, the dependent function
u is prescribed on the boundary of the bounded domain. For example, if the
bounded domain is the rectangular plate 0 < x < L
1
and 0 < y < L
2
, the
56 INTRODUCTION TO PDES
boundary conditions u(0, y), u(L
1
, y), u(x, 0), and u(x, L
2
) are prescribed.
The boundary conditions are called homogeneous if the dependent variable
is zero at any point on the boundary, otherwise the boundary conditions are
called nonhomogeneous.
2. Neumann Boundary Conditions: In this case, ﬁrst partial derivatives
are prescribed on the boundary of the bounded domain. For example, the
Neuman boundary conditions for a rod of length L, where 0 < x < L, are of
the form u
x
(0, t) = α and u
x
(L, t) = β, where α and β are constants.
3. Robin or mixed Boundary Conditions: This occurs when the depen-
dent variable and its ﬁrst partial derivatives are prescribed on the boundary
of the bounded domain.
An initial value problem (or Cauchy problem) is a partial diﬀerential
equation together with a set of additional conditions on the unknwon func-
tion or its derivatives at a point in the given domain of the solution. These
conditions are called initial value conditions. For example, the transport
equation
u
t
(x, t) + cu
x
(x, t) =0
u(x, 0) =f(x).
It can be shown that initial conditions for a pde are necessary and suﬃcient
for the existence of a unique solution.
We say that an initial and/or boundary value problem associated with a PDE
is well-posed if it has a solution which is unique and depends continuously
on the data given in the problem. The last condition, namely the continuous
dependence is important in physical problems. This condition means that
the solution changes by a small amount when the conditions change a little.
Such solutions are said to be stable.
Example 7.4
For x ∈ R and t > 0 we consider the initial value problem
u
tt
−u
xx
=0
u(x, 0) = u
t
(x, 0) =0
Clearly, u(x, t) = 0 is a solution to this problem.
(a) Let 0 < << 1 be a very small number. Show that the function u

(x, t) =
7 SOLUTIONS AND RELATED TOPICS 57

2
sin
_
x

_
sin
_
t

_
is a solution to the problem
u
tt
−u
xx
=0
u(x, 0) =0
u
t
(x, 0) = sin
_
x

_
(b) Show that for a ﬁxed t > 0, sup{|u

(x, t) − u(x, t)| : x ∈ R, t > 0} =
2
.
Thus, a small change in the initial data leads to a small change in the solution.
Hence, the initial value problem is well-posed.
Solution.
(a) We have
∂u

∂t
= sin
_
x

_
cos
_
t

_

2
u

∂t
2
=−sin
_
x

_
sin
_
t

_
∂u

∂x
= cos
_
x

_
sin
_
t

_

2
u

∂x
2
=−sin
_
x

_
sin
_
t

_
Thus,

2
u
∂t
2

2
u
∂x
2
= 0. Moreover, u

(x, 0) = 0 and

∂t
u

(x, 0) = sin
_
x

_
.
(b) We have
sup{|u

(x, t) −u(x, t)| : x ∈ R, t > 0} =
2
sup{
¸
¸
¸
¸
sin
_
x

_
sin
_
t

¸
¸
¸
: x ∈ R, t > 0}
=
2
A problem that is not well-posed is referred to as an ill-posed problem. We
illustrate this concept in the next example.
Example 7.5
For x ∈ R and t > 0 we consider the initial value problem
u
tt
+ u
xx
=0
u(x, 0) = u
t
(x, 0) =0
58 INTRODUCTION TO PDES
Clearly, u(x, t) = 0 is a solution to this problem.
(a) Let 0 < << 1 be a very small number. Show that the function u

(x, t) =

2
sin
_
x

_
sinh
_
t

_
, where
sinh x =
e
x
−e
−x
2
is a solution to the problem
u
tt
+ u
xx
=0
u(x, 0) =0
u
t
(x, 0) = sin
_
x

_
(b) Show that sup{|

∂t
u

(x, 0) − u
t
(x, 0)| : x ∈ R} = and sup{|u

(x, t) −
u(x, t)| : x ∈ R} =
2
¸
¸
sinh
_
t

¸
.
(c) Find lim
t→∞
sup{|u

(x, t) −u(x, t)| : x ∈ R}.
Solution.
(a) We have
∂u

∂t
= sin
_
x

_
cosh
_
t

_

2
u

∂t
2
=sin
_
x

_
sinh
_
t

_
∂u

∂x
= cos
_
x

_
sinh
_
t

_

2
u

∂x
2
=−sin
_
x

_
sinh
_
t

_
Thus,

2
u
∂t
2
+

2
u
∂x
2
= 0. Moreover, u

(x, 0) = 0 and

∂t
u

(x, 0) = sin
_
x

_
.
(b) We have
sup{|

∂t
u

(x, 0) −u
t
(x, 0)| : x ∈ R} =sup{
¸
¸
¸ sin
_
x

¸
¸ : x ∈ R}
= sup{
¸
¸
¸sin
_
x

¸
¸ : x ∈ R} =
and
sup{|u

(x, t) −u(x, t)| : x ∈ R} =
2
sup{
¸
¸
¸
¸
sinh
_
t

_
sin
_
x

_
¸
¸
¸
¸
: x ∈ R}
=
2
¸
¸
¸
¸
sinh
_
t

¸
¸
¸
7 SOLUTIONS AND RELATED TOPICS 59
(c) We have
lim
t→∞
sup{|u

(x, t) −u(x, t)| : x ∈ R} = lim
t→∞

2
¸
¸
¸
¸
sinh
_
t

¸
¸
¸
= ∞.
Thus, a small change in the initial data leads to a catastrophically change in
the solution. Hence, the given problem is ill-posed
60 INTRODUCTION TO PDES
Practice Problems
Exercise 7.1
Determine a and b so that u(x, y) = e
ax+by
is a solution to the equation
u
xxxx
+ u
yyyy
+ 2u
xxyy
= 0.
Exercise 7.2
Consider the following diﬀerential equation
tu
xx
−u
t
= 0.
Suppose u(t, x) = X(x)T(t). Show that there is a constant λ such that
X

= λX and T

= λtT.
Exercise 7.3
Consider the initial value problem
xu
x
+ (x + 1)yu
y
= 0, x, y > 1
u(1, 1) = e.
Show that u(x, y) =
xe
x
y
is the solution to this problem.
Exercise 7.4
Show that u(x, y) = e
−2y
sin (x −y) is the solution to the initial value prob-
lem
_
u
x
+ u
y
+ 2u = 0 for x, y > 1
u(x, 0) = sin x
Exercise 7.5
Solve each of the following diﬀerential equations:
(a)
du
dx
= 0 where u = u(x).
(b)
∂u
∂x
= 0 where u = u(x, y).
Exercise 7.6
Solve each of the following diﬀerential equations:
(a)
d
2
u
dx
2
= 0 where u = u(x).
(b)

2
u
∂x∂y
= 0 where u = u(x, y).
7 SOLUTIONS AND RELATED TOPICS 61
Exercise 7.7
Show that u(x, y) = f(y +2x) +xg(y +2x), where f and g are two arbitrary
twice diﬀerentiable functions, satisfy the equation
u
xx
−4u
xy
+ 4u
yy
= 0.
Exercise 7.8
Find the diﬀerential equation whose general solution is given by u(x, t) =
f(x−ct)+g(x+ct), where f and g are arbitrary twice diﬀerentiable functions
in one variable.
Exercise 7.9
Let p : R →R be a diﬀerentiable function in one variable. Prove that
u
t
= p(u)u
x
has a solution satisfying u(x, t) = f(x + p(u)t), where f is an arbitrary
diﬀerentiable function. Then ﬁnd the general solution to u
t
= (sin u)u
x
.
Exercise 7.10
Find the general solution to the pde
u
xx
+ 2u
xy
+ u
yy
= 0.
Hint: See Exercise 6.2.
62 INTRODUCTION TO PDES
Sample Exam Questions
Exercise 7.11
Let u(x, t) be a function such that u
xx
exists and u(0, t) = u(L, t) = 0 for all
t ∈ R. Prove that
_
L
0
u
xx
(x, t)u(x, t)dx ≤ 0.
Exercise 7.12
Consider the initial value problem
u
t
+ u
xx
= 0, x ∈ R, t > 0
u(x, 0) = 1.
(a) Show that u(x, t) ≡ 1 is a solution to this problem.
(b) Show that u
n
(x, t) = 1 +
e
n
2
t
n
sin nx is a solution to the initial value
problem
u
t
+ u
xx
= 0, x ∈ R, t > 0
u(x, 0) = 1 +
sin nx
n
.
(c) Find sup{|u
n
(x, 0) −1| : x ∈ R}.
(d) Find sup{|u
n
(x, t) −1| : x ∈ R}.
(e) Show that the problem is ill-posed.
Exercise 7.13
Find the general solution of each of the following PDEs by means of direct
integration.
(a) u
x
= 3x
2
+ y
2
, u = u(x, y).
(b) u
xy
= x
2
y, u = u(x, y).
(c) u
xyz
= 0, u = u(x, y, z).
(d) u
xtt
= e
2x+3t
, u = u(x, t).
Exercise 7.14
Consider the second-order PDE
u
xx
+ 4u
xy
+ 4u
yy
= 0.
(a) Use the change of variables v(x, y) = y − 2x and w(x, y) = x to show
that u
ww
= 0.
(b) Find the general solution to the given PDE.
7 SOLUTIONS AND RELATED TOPICS 63
Exercise 7.15
Derive the general solution to the PDE
u
tt
= c
2
u
xx
by using the change of variables v = x + ct and w = x −ct.
64 INTRODUCTION TO PDES
First Order Partial Diﬀerential
Equations
Many problems in the mathematical, physical, and engineering sciences deal
with the formulation and the solution of ﬁrst order partial diﬀerential equa-
tions. Our ﬁrst task is to understand simple ﬁrst order equations. In ap-
plications, ﬁrst order partial diﬀerential equations are most commonly used
to describe dynamical processes, and so time, t, is one of the independent
variables. Most of our discussion will focus on dynamical models in a single
space dimension, bearing in mind that most of the methods can be readily
extended to higher dimensional situations. First order partial diﬀerential
equations and systems model a wide variety of wave phenomena, including
transport of solvents in ﬂuids, ﬂood waves, acoustics, gas dynamics, glacier
motion, traﬃc ﬂow, and also a variety of biological and ecological systems.
From a mathematical point of view, ﬁrst order partial diﬀerential equations
have the advantage of providing conceptual basis that can be utilized in the
study of higher order partial diﬀerential equations.
In this chapter we introduce the basic deﬁnitions of ﬁrst order partial dif-
ferential equations. We then derive the one dimensional spatial transport
eqution and discuss some methods of solutions. One general method of solv-
ability for quasilinear ﬁrst order partial diﬀerential equation, known as the
method of characteristics, is analyzed.
8 Classiﬁcation of First Order PDEs
In this section, we present the basic deﬁnitions pertained to ﬁrst order PDE.
By a ﬁrst order diﬀerential equation in two variables x and y we mean
65
66 FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS
any equation of the form
F(x, y, u, u
x
, u
y
) = 0. (8.1)
In what follows the functions a, b, and c are assumed to be continuously
diﬀerentiable functions. If Equation (8.1) can be written in the form
a(x, y, u)u
x
+ b(x, y, u)u
y
= c(x, y, u) (8.2)
then we say that the equation is quasilinear. The following are examples
of quasilinear equations:
uu
x
+ u
y
+ cu
2
= 0
x(y
2
+ u)u
x
−y(x
2
+ u)u
y
= (x
2
−y
2
)u.
If Equation (8.1) can be written in the form
a(x, y)u
x
+ b(x, y)u
y
= c(x, y, u) (8.3)
then we say that the equation is semilinear. The following are examples of
semilinear equations:
xu
x
+ yu
y
= u
2
+ x
2
(x + 1)
2
u
x
+ (y −1)
2
u
y
= (x + y)u
2
.
If Equation (8.1) can be written in the form
a(x, y)u
x
+ b(x, y)u
y
+ c(x, y)u = d(x, y) (8.4)
then we say that the equation is linear. Examples of linear equations are:
xu
x
+ yu
y
= cu
(y −z)y
x
+ (z −x)u
y
+ (x −y)u
z
= 0.
A ﬁrst order pde that is not linear is said to be nonlinear. Examples of
nonlinear equations are:
u
x
+ cu
2
y
= xy
u
2
x
+ u
2
y
= c.
First order partial diﬀerential equations are classiﬁed as either linear or non-
linear. Clearly, linear equations are a special kind of quasilinear equation
8 CLASSIFICATION OF FIRST ORDER PDES 67
(8.2) if a and b are functions of x and y only and c is a linear function of u.
Likewise, semilinear equations are quasilinear equations if a and b are func-
tions of x and y only. Also, semilinear equations (8.4) reduces to a linear
equation if c is linear in u.
A linear equation is called homogeneous if d(x, y) ≡ 0 and nonhomoge-
neous if d(x, y) = 0. Examples of linear homogeneous equations are:
xu
x
+ yu
y
= cu
(y −z)y
x
+ (z −x)u
y
+ (x −y)u
z
= 0.
Examples of nonhomogeneous equations are:
u
x
+ (x + y)u
y
−u = e
x
yu
x
+ xu
y
= xy.
Recall that for an ordinary linear diﬀerential equation, the general solution
depends mainly on arbitrary constants. Unlike ODEs, in linear partial dif-
ferential equations, the general solution depends on arbitrary functions.
Example 8.1
Solve the equation u
t
(t, x) = 0.
Solution.
The general solution is given by u(t, x) = f(x) where f is an arbitrary dif-
ferentiable function of x
Example 8.2
Consider the transport equation
au
t
(t, x) + bu
x
(t, x) = 0
where a and b are constants. Show that u(t, x) = f(bt − ax) is a solution
to the given equation, where f is an arbitrary diﬀerentiable function in one
variable.
Solution.
Let v(t, x) = bt −ax. Using the chain rule we see that u
t
(t, x) = bf
v
(v) and
u
x
(t, x) = −af
v
(v). Hence, au
t
(t, x) + bu
x
(t, x) = abf
v
(v) −abf
v
(v) = 0
68 FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS
Practice Problems
Exercise 8.1
Classify each of the following PDE as linear, quasilinear, semi-linear, or non-
linear.
(a) xu
x
+ yu
y
= sin (xy).
(b) u
t
+ uu
x
= 0
(c) u
2
x
+ u
3
u
4
y
= 0.
(d) (x + 3)u
x
+ xy
2
u
y
= u
3
.
Exercise 8.2
Show that u(x, y) = e
x
f(2x −y), where f is a diﬀerentiable function of one
variable, is a solution to the equation
u
x
+ 2u
y
−u = 0.
Exercise 8.3
Show that u(x, y) = x

xy satisﬁes the equation
xu
x
−yu
y
= u
subject to
u(y, y) = y
2
.
Exercise 8.4
Show that u(x, y) = cos (x
2
+ y
2
) satisﬁes the equation
−yu
x
+ xu
y
= 0
subject to
u(0, y) = cos y
2
.
Exercise 8.5
Show that u(x, y) = y −
1
2
(x
2
−y
2
) satisﬁes the equation
1
x
u
x
+
1
y
u
y
=
1
y
subject to u(x, 1) =
1
2
(3 −x
2
).
8 CLASSIFICATION OF FIRST ORDER PDES 69
Exercise 8.6
Find a relationship between a and b if u(x, y) = f(ax + by) is a solution to
the equation 3u
x
−7u
y
= 0 for any diﬀerentiable function f.
Exercise 8.7
Suppose L is a linear operator, that is, L(αu+βv) = αL(u)+βL(v). Consider
the homogeneous and inhomogeneous linear equations
Lu = 0
Lu = f
where f is some function. Suppose v is a solution to the homogeneous equa-
tion, and w is a solution to the inhomogeneous equation. Show u = av + w
is a solution to the inhomogeneous equation for any constant a.
Exercise 8.8
Reduce the partial diﬀerential equation
au
x
+ bu
y
+ cu = 0
to a ﬁrst order ODE by introducing the change of variables s = ax +by and
t = bx −ay.
Exercise 8.9
Solve the partial diﬀerential equation
u
x
+ u
y
= 1
by introducing the change of variables s = x + y and t = x −y.
Exercise 8.10
Consider the IVP
−xu
x
+ u
t
= 0, u(x, 0) = f(x) ≥ 0.
Suppose that
lim
x→∞
xf(x) = 0.
We will see later in the book that the solution to this problem is a nonnegative
function. Deﬁne I(t) =
_

−∞
u(x, t)dx. Show that I(t) is a decreasing function
of t.
70 FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS
Sample Exam Questions
Exercise 8.11
Show that u(x, y) = e
−4x
f(2x −3y) is a solution to the ﬁrst-order PDE
3u
x
+ 2u
y
+ 12u = 0.
Exercise 8.12
Derive the general solution of the PDE
au
t
+ bu
x
= u, a, b = 0
by using the change of variables v = ax −bt and w =
1
a
t.
Exercise 8.13
Derive the general solution of the PDE
au
t
+ bu
x
= 0, a, b = 0
by using the change of variables s(x, y) = ax + by and t(x, y) = bx − ay.
Assume a
2
+ b
2
> 0.
Exercise 8.14
Write the equation
u
t
+ cu
x
+ λu = f(x, y)
in the coordinates v = x −ct, w = t.
Exercise 8.15
Suppose that u(x, t) = w(x −ct) is a solution to the PDE
xu
x
+ tu
t
= Au
where A and c are constants. Let v = x −ct. Write the diﬀerential equation
with unknown function w(v).
9 THE ONE DIMENSIONAL SPATIAL TRANSPORT EQUATIONS 71
9 The One Dimensional Spatial Transport Equa-
tions
Modeling is the process of writing a diﬀerential equation to describe a physi-
cal situation. In this section we discuss the one-dimensional transport equa-
tion and discuss an analytical method for solving it.
Linear Transport Equation for Fluid Flows
We shall describe the transport of a dissolved chemical by water that is trav-
eling with uniform velocity c through a long thin tube G with uniform cross
section A. (The very same discussion applies to the description of the trans-
port of gas by air moving through a pipe.) We identify G with the open
interval (a, b), and the velocity c > 0 is in the (rightward) positive direction
of the x−axis. We will assume that the concentration of the chemical is con-
stant across the cross section A at each point x so that the chemical changes
in the x−direction and thus the term one-dimensional spatial equation. See
Figure 9.1
Figure 9.1
Let u(x, t) be a continuously diﬀerentiable function denoting the concentra-
tion of the chemical (i.e. amount of chemical/area) at position x at time
t. Then at time t, the amount of chemical stored in a section of the tube
between positions a and x is given by the deﬁnite integral
_
x
a
Au(s, t)ds.
72 FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS
Since the water is ﬂowing at a speed c, so at time h +t the same quantity of
chemical will be
_
x
a
Au(s, t)ds =
_
x+ch
a+ch
u(s, t + h)ds.
Taking the derivative of both sides with respec to to x we ﬁnd
u(x, t) = u(x + ch, t + h).
Now taking the derivative of this last equation we respect to h we ﬁnd
0 = u
t
(x + ch, t + h) + cu
x
(x + h, t + h).
Taking the limit of this last equation as h approaches 0 we ﬁnd
u
t
(x, t) + cu
x
(x, t) = 0 (9.1)
for all (x, t). This equation is called the transport equation in one-dimensional
space. It is a linear, homogeneous ﬁrst order partial diﬀerential equation.
Example 9.1
Show that u(x, t) = f(x − ct) is a solution to (9.1), where f is an arbitrary
diﬀerentiable function in one variable.
Solution.
Using the chain rule we ﬁnd
u
t
= −cf

(x −ct) and u
x
= f

(x −ct).
Hence, by substituting these results into the equation we ﬁnd
u
t
+ cu
x
= −cf

(x −ct) + cf

(x −ct) = 0
The solution u(x, t) = f(x − ct) is called the right traveling wave, since
the graph of the function f(x − ct) at a given time t is the graph of f(x)
shifted to the right by the value ct. Thus, with growing time, the function
f(x) is moving without changes to the right at the speed c
An initial value condition determines a unique solution to the transport equa-
tion as stated in the next theorem.
9 THE ONE DIMENSIONAL SPATIAL TRANSPORT EQUATIONS 73
Theorem 9.4
Let f be a continuously diﬀerentiable function. Then there is a unique con-
tinuously diﬀerentiable solution u(x, t) to the IVP
au
t
(x, t) + bu
x
(x, t) = 0
u(x, 0) = f(x).
Indeed, u is given explicitly by the formula
u(x, t) = f(bt −ax).
Method of Solutions: The Coordinate Method
We will solve (9.1) by solving the more general equation
au
x
+ bu
y
= 0 (9.2)
where a
2
+ b
2
> 0.
We introduce a new rectangular system by the substitution
s = ax + by, t = bx −ay
According to the chain rule for the derivative of a composite function, we
have
u
x
=u
s
s
x
+ u
t
t
x
= au
s
+ bu
t
u
y
=u
s
s
y
+ u
t
t
y
= bu
s
−au
t
Substituting these into (9.2) to obtain
a
2
u
s
+ abu
t
+ b
2
u
s
−abu
t
= 0
or
(a
2
+ b
2
)u
s
= 0
and since a
2
+ b
2
> 0 we obtain
u
s
= 0.
Solving this equation, we ﬁnd
u(s, t) = f(w)
where f is an arbitrary diﬀerentiable function of one variable. Now, in terms
of x and y we ﬁnd
u(x, y) = f(bx −ay).
74 FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS
Example 9.2
Use the coordinate method to ﬁnd the solution to u
t
−3u
x
= 0, u(x, 0) = e
−x
2
.
Solution.
Let v = −3x + t and w = x + 3t. Then u
x
= −3u
v
+ u
w
and u
t
= u
v
+ 3u
w
.
Substituting these into the given equation we ﬁnd 10u
v
= 0 or u
v
= 0.
Hence, u(v, w) = f(w) or u(x, t) = f(x + 3t) where f is a diﬀerentiable
function in one variable. Since u(x, 0) = e
−x
2
, we ﬁnd e
−x
2
= f(x). Hence,
u(x, t) = e
−(x+3t)
2
Transport Equation with Decay: The Method of Characteristic
Coordinates
Recall from ODE that a function u is an exponential growth function if it
satisﬁes the equation
du
dt
= λu.
A transport equation with growth is an equation given by
u
t
+ cu
x
+ λu = f(x, t)
where λ and c are constants and f is a given function representing external
resources. Note that the growth is characterized by the term λu.
To solve this equation, we introduce the characteristic coordinates given
by
v = x −ct, w = t.
Using the chain rule, we ﬁnd
u
t
=u
v
v
t
+ u
w
w
t
= −cu
v
+ u
w
u
x
=u
v
v
x
+ u
w
w
x
= u
v
Substituting these into the original equation we obtain the equation
u
w
+ λu = f(v + cw, w)
which can be solved by the method of integrating factor. We illustrate this
approach in the next example.
Example 9.3
Find the general solution of the transport equation
u
t
+ u
x
−u = t.
9 THE ONE DIMENSIONAL SPATIAL TRANSPORT EQUATIONS 75
Solution.
The characteristic coordinates are
v = x −t, w = t.
These transform the original equation to the ﬁrst order ODE
u
w
−u = w.
Using the method of integrating factor, we ﬁnd
d
dw
(e
−w
u) = we
−w
and solving this equation we ﬁnd
u(v, w) = −(1 + w) + e
w
f(v)
and in terms of x and y we ﬁnd
u(x, t) = f(x −t)e
t
−(1 + t)
A more general method for solving quasilinear ﬁrst order partial diﬀerential
equations, known as the method of characteristics, will be discussed in the
next section.
76 FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS
Practice Problems
Exercise 9.1
Use the coordinate method to ﬁnd the solution to u
t
+ 3u
x
= 0, u(x, 0) =
sin x.
Exercise 9.2
Use the coordinate method, solve the equation au
x
+ bu
y
+ cu = 0.
Exercise 9.3
Use the coordinate method, solve the equation u
x
+ 2u
y
= cos (y −2x) with
the initial condition u(0, y) = f(y), where f : R →R is a given function.
Exercise 9.4
Show that the initial value problem u
t
+u
x
= x, u(x, x) = 1 has no solution.
Exercise 9.5
Solve the transport equation u
t
+2u
x
= −3u with initial condition u(x, 0) =
1
1+x
2
.
Exercise 9.6
Solve u
t
+ u
x
−3u = t with initial condition u(x, 0) = x
2
.
Exercise 9.7
Show that the decay term λu in the transport equation with decay
u
t
+ cu
x
+ λu = 0
can be eliminated by the substitution w = ue
λt
.
Exercise 9.8
Use the coordinate method to solve
u
x
+ u
y
= u
2
u(x, 0) = h(x)
9 THE ONE DIMENSIONAL SPATIAL TRANSPORT EQUATIONS 77
Exercise 9.9 (Well-Posed)
Let u be the unique solution to the IVP
u
t
+ cu
x
= 0
u(x, 0) = f(x)
and v be the unique solution to the IVP
u
t
+ cu
x
= 0
u(x, 0) = g(x)
where f and g are continuously diﬀerentiable functions.
(a) Show that w(x, t) = u(x, t) −v(x, t) is the unique solution to the IVP
u
t
+ cu
x
= 0
u(x, 0) = f(x) −g(x)
(b) Write an explicit formula for w in terms of f and g.
(c) Use (b) to conclude that the transport problem is well-posed. That is, a
small change in the initial data leads to a small change in the solution.
Exercise 9.10
Solve the initial boundary value problem
u
t
+ cu
x
= −λu, x > 0, t > 0
u(x, 0) = 0, u(0, t) = g(t), t > 0.
78 FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS
Sample Exam Questions
Exercise 9.11
Solve the ﬁrst-order equation 2u
t
+3u
x
= 0 with the initial condition u(x, 0) =
sin x.
Exercise 9.12
Solve the PDE
u
x
+ u
y
= 1
using the coordinate method.
Exercise 9.13
Consider the ﬁrst order linear homogeneous PDE
Au
x
+ Bu
y
+ Cu = 0
where A, B, and C are constants with A = 0.
(a) Determine a, b, c, d in terms of A, B, C such that ad −bc = 0 and so that
the change of variables v = ax + by and w = cx + dy will reduce the give
PDE to a ﬁrst order PDE of the form αu
v
+ βu = 0.
(b) Use (a) to ﬁnd the general solution of the given PDE.
Exercise 9.14
Use the result of the previous section to solve the PDE
u
x
+ u
y
+ u = 0.
Exercise 9.15
Consider the ﬁrst order linear nonhomogeneous PDE
Au
x
+ Bu
y
+ Cu = G(x, y)
where A, B, and C are constants with A = 0 and G(x, y) an arbitrary func-
tion. As in the case of ODEs, the general solution to this equation is the sum
of the general solution of the homogeneous equation and a particular solu-
tion of the nonhomogeneous equation. A fairly general method iin ﬁnding a
particular solution is to use an integrating factor, exactly as with ODEs, by
writing the Au
v
+Cu = G(x, y) based on the formulas v = x and w = Bx−Ay
in the form

∂v
(Aue
C
A
v
) = e
C
A
v
G
_
v,
1
A
(Bv −w)
_
.
9 THE ONE DIMENSIONAL SPATIAL TRANSPORT EQUATIONS 79
The right-hand side may be integrated with respect to v, treating w as a
constant.
Using the above idea, ﬁnd the general solution of the PDE
u
x
+ u
y
+ u = x + y.
80 FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS
10 The Method of Characteristics
In this section we develop a method for ﬁnding the general solution of a
quasilinear ﬁrst order partial diﬀerential equation. This method is called
the method of characteristics or Lagrange’s method. This method of
solution can be described by the following result.
Theorem 10.1
The general solution of the quasilinear ﬁrst order PDE
a(x, y, u)u
x
+ b(x, y, u)u
y
= c(x, y, u) (10.1)
is given by
f(v, w) = 0 (10.2)
where f is an arbitrary diﬀerentiable function of v(x, y, u) and w(x, y, u) and
v =constant= c
1
, w =constant= c
2
are solutions to the ODE system
dx
a
=
dy
b
=
du
c
. (10.3)
Equations (10.3) are called the characteristic equations in non-parametric
forms. The corresponding parametric forms are given by the system of ODEs
dx
ds
=a
dy
ds
=b
du
ds
=c
Remark 10.1
Sometimes (10.2) is written explicitly as v = g(w) or w = g(v) where g is an
arbitrary diﬀerentiable function.
Example 10.1
Find the general solution of the pde x
2
u
x
+ y
2
u
y
= (x + y)u.
Solution.
The characteristic equations for this pde are
dx
x
2
=
dy
y
2
=
du
(x+y)u
. Using the ﬁrst
10 THE METHOD OF CHARACTERISTICS 81
two fractions, we have
_
dx
x
2
=
_
dy
y
2
and this implies
x−y
xy
= c
1
. Also, we can
solve for x obtaining x =
1
1
y
−c
1
and x+y =
y
1−c
1
y
+y. Using the last two frac-
tions we ﬁnd
dy
y
2
=
du
u(x+y
=⇒
1
y
2
_
y
1−c
1
y
+ y
_
dy =
du
u
=⇒
_
y−c
1
y
2
+
1
y
_
dy =
du
u
=⇒
_
1−c
1
y
y−c
1
y
2
+
c
1
1−c
1
y
+
1
y
_
dy =
du
u
=⇒
_
_
2
y
+
c
1
1−c
1
y
_
=
_
du
u
=⇒ 2 ln y −
ln (1 −c
1
y) = ln z + c
2
=⇒
y
2
1−c
1
y
= e
c
2
z =⇒ y ·
y
1−c
1
y
= c

2
z =⇒ y
1
1
y
−c
1
=
c

2
z =⇒xy = c

2
z. Hence, the general solution is
f
_
x −y
xy
,
xy
u
_
= 0
where f is an arbitrary diﬀerentiable function
Example 10.2
Find the general solution of the pde yuu
x
+ xuu
y
= xy.
Solution.
The characteristic equations are
dx
yu
=
dy
xu
=
du
xy
. Using the ﬁrst two fractions
we ﬁnd x
2
−y
2
= c
1
. Using the last two fractions we ﬁnd u
2
−y
2
= c
2
. Hence,
the general solution is f(x
2
−y
2
, u
2
−y
2
) = 0 or u
2
= y
2
+g(x
2
−y
2
), where
f and g are arbitrary diﬀerentiable functions
Example 10.3
Find the general solution of the pde x(y
2
−u
2
)u
x
−y(u
2
+x
2
)y
y
= (x
2
+y
2
)u.
Solution.
The characteristic equations are
dx
x(y
2
−u
2
)
=
dy
−y(u
2
+x
2
)
=
du
(x
2
+y
2
)u
. Using a
property of proportions we can write
xdx + ydy + udu
x
2
(y
2
−u
2
) −y
2
(u
2
+ x
2
) + u
2
(x
2
+ y
2
=
du
(x
2
+ y
2
)u
.
That is
xdx + ydy + udu
0
=
du
(x
2
+ y
2
)u
or
xdx + ydy + udu = 0.
82 FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS
Hence, we ﬁnd x
2
+ y
2
+ u
2
= c
1
. Also,
dx
x

dy
y
y
2
−u
2
+ u
2
+ x
2
=
du
(x
2
+ y
2
)u
or
dx
x

dy
y
=
du
u
.
Hence, we ﬁnd ln
_
yu
x
_
=constant or
yu
x
= c
2
. The general solution is given
by
f
_
x
2
+ y
2
+ u
2
,
yu
x
_
= 0
or
u =
x
y
g(x
2
+ y
2
+ u
2
)
where f and g are arbitrary diﬀerentiable functions
Example 10.4
Solve the transport equation using the method of characteristics
u
t
+ cu
x
= 0.
Solution.
The characteristic equations are given by
dt
1
=
dx
c
=
du
0
.
Solving the ﬁrst two fractions we ﬁnd x − ct = k. The last fraction implies
u = k

. The general solution is given by f(x −ct, u) = 0 or u = g(x −ct)
Solution curves to the ODE
dy
dx
=
b
a
are called characteristic curves or simply characteristics. These are
curves in the xy−plane.
Example 10.5
Find the characteristics of cos yu
x
+ u
y
+ xu = 0.
10 THE METHOD OF CHARACTERISTICS 83
Solution.
Solving the equation
dy
dx
=
1
cos y
by the separation of variable method we ﬁnd
sin y −x = k
Example 10.6
Find the characteristics of u
x
+ 2u
y
−u = 0.
Solution.
We have a = 1 and b = 2. Thus,
dy
dx
= 2 so that the characteristics are given
by 2x −y = k
84 FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS
Practice Problem
Exercise 10.1
Find the characteristics of the pde
xu
x
−yu
y
= u
Exercise 10.2
Find the characteristics of the pde
−yu
x
+ xu
y
= 0
Exercise 10.3
Find the characteristics of the pde
(x + y)(u
x
+ u
y
) = u −1
Exercise 10.4
Find the general solution of the pde xu
x
+ yu
y
= 1 + u
2
.
Exercise 10.5
Find the general solution of the pde ln (y + u)u
x
+ u
y
= −1.
Exercise 10.6
Find the general solution of the pde xu
x
+ yu
y
= u.
Exercise 10.7
Find the general solution of the pde xu
x
+ yu
y
= nu.
Exercise 10.8
Find the general solution of the pde x(y −u)u
x
+ y(u −x)u
y
= u(x −y).
Exercise 10.9
Find the general solution of the pde u(u
2
+ xy)(xu
x
−yu
y
) = x
4
.
Exercise 10.10
Find the general solution of the pde (y + xu)u
x
−(x + yu)u
y
= x
2
−y
2
.
Exercise 10.11
Find the general solution of the pde (y
2
+ u
2
)u
x
−xyu
y
+ xu = 0.
10 THE METHOD OF CHARACTERISTICS 85
Exercise 10.12
Find the general form of solutions to
u
x
+ 2u
y
= u
and sketch some of the characteristics. Hint: deﬁne a new variable v = e
−x
u.
What equation does v satisfy?
Exercise 10.13
Find the general form of solutions to
(1 + x
2
)u
x
+ u
y
= 0
and sketch some of the characteristics.
86 FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS
Sample Exam Questions
Exercise 10.14
Find the general solution of the equation
u
x
+ yu
y
= u.
Exercise 10.15
Find the characteristics associated with the PDE
u
x
+ xu
y
+ 3u = 2.
Exercise 10.16
Find the general solution of hte ﬁrst order PDE
u
x
+ yu
y
+ xu = 0.
Exercise 10.17
Find the characteristics of the pde
1
x
u
x
+
1
y
u
y
= 0
Exercise 10.18
Find the characteristics of the pde
1
x
u
x
+
1
y
u
y
=
1
y
11 THE CAUCHYPROBLEMFOR FIRST ORDER QUASILINEAR EQUATIONS87
11 The Cauchy Problem for First Order Quasi-
linear Equations
When solving a partial diﬀerential equation, it is seldom the case that one
tries to study the properties of the general solution of such equations. In
general, one deals with those partial diﬀerential equations whose solutions
satisfy certain supplementary conditions. In the case of a ﬁrst order partial
diﬀerential equation, we determine the particular solution by formulating an
initial value porblem also known as a Cauchy problem.
In this section, we discuss the Cauchy problem for the ﬁrst order quasilinear
partial diﬀerential equation
a(x, y, u)u
x
+ b(x, y, u)u
y
= c(x, y, u). (11.1)
Recall that the initial value problem of a ﬁrst order ordinary diﬀerential
equation asks for a solution of the equation which has a given value at a
given point in R. The Cauchy problem for the PDE (11.1) asks for a solution
of (11.1) which has given values on a given curve in R
2
. A precise statement
of the problem is given next.
Initial Value Problem or Cauchy Problem
Let C be a given curve in R
2
deﬁned parametrically by the equations
x = x
0
(t), y = y
0
(t)
where x
0
, y
0
are continuously diﬀerentiable functions on some interval I. Let
u
0
(t) be a given continuously diﬀerentiable function on I. The Cauchy prob-
lem for (11.1) asks for a continuously diﬀerentiable function u = u(x, y)
deﬁned in a domain Ω ⊂ R
2
containing the curve C and such that:
(1) u = u(x, y) is a solution of (11.1) in Ω.
(2) On the curve C, u equals the given function u
0
(t), i.e.
u(x
0
(t), y
0
(t)) = u
0
(t), t ∈ I. (11.2)
We call C the initial curve of the problem, u
0
(t) the initial data, and
(11.2) the initial condition of the problem. See Figure 11.1.
88 FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS
Figure 11.1
If we view a solution u = u(x, y) of (11.1) as an integral surface of (11.1),
we can give a simple geometrical statement of the problem: Find a solu-
tion surface of (11.1) containing the curve Γ described parametrically by the
equations
Γ : x = x
0
(t), y = y
0
(t), u = u
0
(t), t ∈ I.
Note that the projection of this curve in the xy−plane is just the curve C.
The following theorem asserts that under certain conditions the Cauchy prob-
lem (11.1) - (11.2) has a unique solution.
Theorem 11.1
Suppose that x
0
(t), y
0
(t), and u
0
(t) are continuously diﬀerentiable functions
of t in an interval I, and that a, b, and c are functions of x, y, and u with
continuous ﬁrst order partial derivatives with respect to their argument in
some domain D of (x, y, u)−space containing the initial curve
Γ : x = x
0
(t), y = y
0
(t), u = u
0
(t)
where t ∈ I. Then for each point (x
0
(t), y
0
(t), u
0
(t)) on Γ that satisﬁes the
condition
a(x
0
(t), y
0
(t), u
0
(t))
dy
0
dt
(t) −b(x
0
(t), y
0
(t), u
0
(t))
dx
0
dt
(t) = 0. (11.3)
there exists a unique solution u = u(x, y) of (11.1) in a neighborhood U of
(x
0
(t), y
0
(t)) such that the initial condition (11.2) is satisﬁed for every point
on C contained in U. See Figure 11.2.
11 THE CAUCHYPROBLEMFOR FIRST ORDER QUASILINEAR EQUATIONS89
Figure 11.2
Note that condition (11.3) implies that
dy
0
(t)
dx
0
(t)
=
b(x
0
, y
0
, u
0
)
a(x
0
, y
0
, u
0
)
which means that the vector (a(x
0
, y
0
, u
0
), b(x
0
, y
0
, u
0
), c(x
0
, y
0
, u
0
)) is not
tangent to Γ. It follows that the Cauchy problem has a unique solution if C
is nowhere characteristic.
We construct the desired solution using the method of characteristics as
follows: Pick a point (x
0
(t), y
0
(t), u
0
(t)) ∈ Γ. Using this as the initial value
we solve the system of ODEs consisting of the characteristic equations in
parametric form
dx
ds
=a
dy
ds
=b
du
ds
=c
satisfying the initial condition
(x(0), y(0), u(0)) = (x
0
(t), y
0
(t), u
0
(t)).
The solution depends on the parameter s so it consists of a triples of functions
x = x(s, t), y = y(s, t), u = u(s, t) (11.4)
90 FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS
This system represents the parametric representation of the integral surface
of the problem in which the curve Γ corresponds to s = 0. The solution u is
recovered by solving the ﬁrst two equations in (11.4) for
t = t(x, y), s = s(x, y)
and substituting these into the third equation to obtain u(x, y) = u(s(x, y), t(x, y)).
Example 11.1
Solve the Cauchy problem
u
x
+ u
y
=1
u(x, 0) =f(x)
Solution.
The initial curve in R
3
can be given parametrically as
Γ : x
0
(t) = t, y
0
(t) = 0, u
0
(t) = f(t).
We have
a(x
0
(t), y
0
(t), u
0
(t))
dy
0
dt
(t) −b(x
0
(t), y
0
(t), u
0
(t))
dx
0
dt
(t) = −1 = 0
so by the above theorem the given Cauchy problem has a unique solution.
To ﬁnd this solution, we solve the system of ODEs
dx
ds
=1
dy
ds
=1
du
ds
=1.
Solving this system we ﬁnd
x(s) = s + c
1
, y(s) = s + c
2
, u(s) = s + c
3
.
But x(0) = t so that c
1
= t. Similarly, y(0) = 0 so that c
2
= 0 and u(0) = f(t)
implies c
3
= f(t). Hence, the unique solution is given parametrically by the
equations
x(s, t) = t + s, y(s, t) = s, u(s, t) = s + f(t).
11 THE CAUCHYPROBLEMFOR FIRST ORDER QUASILINEAR EQUATIONS91
Solving the ﬁrst two equations for s and t we ﬁnd
s = y, t = x −y
and substituting these into the third equation we ﬁnd
u(x, y) = y + f(x −y).
Alternative Computation
We can apply the results of the previous section to ﬁnd the unique solution.
If we solve the characteristic equations in non-parametric form
dx
1
=
dy
1
=
du
1
we ﬁnd x − y = c
1
and u − x = c
2
. Thus, the general solution of the PDE
is given by u = x +F(x −y). Using the Cauchy data u(x, 0) = f(x) we ﬁnd
f(x) = x + F(x) which implies that F(x) = f(x) − x. Hence, the unique
solution is given by
u(x, y) = x + f(x −y) −(x −y) = y + f(x −y)
If condition (11.3) is not satisﬁed than C is a characteristic curve. If the
curve Γ satisﬁes the characteristic equations than the problem has inﬁnitely
many solutions. To see this, pick an arbitrary point P
0
= (x
0
, y
0
, u
0
) on Γ.
Pick a new initial curve Γ

passing through P
0
which is not tangent to Γ at
P
0
. In this case, condition (11.3) is satisﬁed and the new Cauchy problem has
a unique solution. Since there are inﬁnitely many ways of selecting Γ

, we
obtain inﬁnitely many solutions. We illustrate this case in the next example.
Example 11.2
Solve the Cauchy problem
u
x
+ u
y
=1
u(x, x) =x
Solution.
The initial curve in R
3
can be given parametrically as
Γ : x
0
(t) = t, y
0
(t) = t, u
0
(t) = t.
92 FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS
We have
a(x
0
(t), y
0
(t), u
0
(t))
dy
0
dt
(t) −b(x
0
(t), y
0
(t), u
0
(t))
dx
0
dt
(t) = 0.
As in Example 11.1, the general solution of the PDE is u = x+f(x−y) where
f is an arbitrary diﬀerentiable function. Using the Cauchy data u(x, x) = x
we ﬁnd f(0) = 0. Thus, the solution is given by
u(x, y) = y + f(x −y)
where f is an arbitrary function such that f(0) = 0. There are inﬁnitely
many choices for f. Hence, the problem has inﬁnitely many solutions. Note
that Γ satisﬁes the characteristic equations
If condition (11.3) is not satisﬁed and if Γ does not satisfy the characteristic
equations then it can be shown that the Cauchy problem has no solutions.
We illustrate this case next.
Example 11.3
Solve the Cauchy problem
u
x
+ u
y
=1
u(x, x) =1
Solution.
The initial curve in R
3
can be given parametrically as
Γ : x
0
(t) = t, y
0
(t) = t, u
0
(t) = 1.
We have
a(x
0
(t), y
0
(t), u
0
(t))
dy
0
dt
(t) −b(x
0
(t), y
0
(t), u
0
(t))
dx
0
dt
(t) = 0.
Solving the characteristic equations in parametric form we ﬁnd
x(s) = s + c
1
, y(s) = s + c
2
, u(s) = s + c
3
.
Clearly, Γ does not satisfy the characteristic equations. Now, the general
solution to the PDE is given by u = x + f(x − y). Using the Cauchy data
u(x, x) = 1 we ﬁnd f(0) = 1 − x, which is not possible since the LHS is a
ﬁxed number whereas the RHS is a variable expression. Hence, the problem
has no solutions
11 THE CAUCHYPROBLEMFOR FIRST ORDER QUASILINEAR EQUATIONS93
Example 11.4
Solve the Cauchy problem
u
x
−u
y
=1
u(x, 0) =x
2
(11.5)
Solution.
The initial curve is given parametrically by
Γ : x
0
(t) = t, y
0
(t) = 0, u
0
(t) = t
2
.
We have
a(x
0
(t), y
0
(t), u
0
(t))
dy
0
dt
(t) −b(x
0
(t), y
0
(t), u
0
(t))
dx
0
dt
(t) = 1 = 0
so the Cauchy problem has a unique solution.
The characteristic equations are
dx
1
=
dy
−1
=
du
1
.
Using the ﬁrst two fractions we ﬁnd x +y = c
1
. Using the ﬁrst and the third
fractions we ﬁnd u − x = c
2
. Thus, the general solution can be represented
by
u −x = f(x + y)
where f is an arbitrary diﬀerentiable function. Using the Cauchy data
u(x, 0) = x
2
we ﬁnd x
2
− x = f(x). Hence, the unique solution is given
by
u = x + (x + y)
2
−(x + y) = (x + y)
2
−y
Example 11.5
Solve the initial value problem
u
t
+ uu
x
= x, u(x, 0) = 1.
Solution.
The initial curve is given parametrically by
Γ : x
0
(s) = s, y
0
(s) = 0, u
0
(s) = 1.
94 FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS
We have
a(x
0
(s), y
0
(s), u
0
(s))
dy
0
ds
(s) −b(x
0
(s), y
0
(s), u
0
(s))
dx
0
ds
(s) = −1 = 0
so the Cauchy problem has a unique solution.
The characteristic equations are
dt
1
=
dx
u
=
du
x
.
Since
dt
1
=
d(x + u)
x + u)
we ﬁnd that t = ln (x + u) −ln c
1
or (x+u)e
−t
= c
1
. Now, using the last two
fractions we ﬁnd u
2
−x
2
= c
2
. Hence, the general solution is givne by
f((x + u)e
−t
, u
2
−x
2
) = 0
where f is an arbitrary diﬀerentiable function. Using the Cauchy data we
ﬁnd c
1
= 1 + x and c
2
= 1 −x
2
= 2(1 + x) −(1 + x)
2
= 2c
1
−c
2
1
. Thus,
u
2
−x
2
= 2(x + u)e
−t
−(x + u)
2
e
−2t
or
u −x = 2e
−t
−(x + u)e
−2t
.
This can be reduced further as follows: u + ue
−2t
= x + 2e
−t
− xe
−2t
=
2e
−t
+ x(1 −e
−2t
) =⇒u =
2e
−t
1+e
−2t
+ x
1−e
−2t
1+e
−2t
= sech(t) + xtanh(t)
Example 11.6
Solve the initial value problem
uu
x
+ u
y
= 1
with the initial curve
Γ : x
0
(s) = 2s
2
, y
0
(s) = 2s, u
0
(s) = 0, s > 0.
Solution.
We have
a(x
0
(s), y
0
(s), u
0
(s))
dy
0
ds
(s) −b(x
0
(s), y
0
(s), u
0
(s))
dx
0
ds
(s) = −4s = 0, s > 0
11 THE CAUCHYPROBLEMFOR FIRST ORDER QUASILINEAR EQUATIONS95
so the Cauchy problem has a unique solution.
The characteristic equations in parametric form are given by the system of
ODEs
dx
dt
=u
dy
dt
=1
du
dt
=1
Thus, the solution of this system depends on two parameters s and t. Solving
the last two equations we ﬁnd
y(s, t) = t + c
2
, u(s, t) = t + c
3
.
Solving the ﬁrst equation with u being replaced by t + c
3
we ﬁnd
x(s, t) =
1
2
t
2
+ c
3
t + c
1
.
Using the initial conditions
x(s, 0) = 2s
2
, y(s, 0) = 2s, u(s, 0) = 0
we ﬁnd
x(s, t) =
1
2
t
2
+ 2s
2
, y(s, t) = t + 2s, u(s, t) = t.
Eliminating s and t we ﬁnd
(u −y)
2
+ u
2
= 2x.
Solving this quadratic equation in u to ﬁnd
2u = y ±(4x −y
2
)
1
2
.
The solution surface satisfying u = 0 on y
2
= 2x is given by
2u = y −(4x −y
2
)
1
2
.
This represents a solution surface only when y
2
< 4x. The solution does not
exist for y
2
> 4x
96 FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS
Practice Problems
Exercise 11.1
Solve
(y −u)u
x
+ (u −x)u
y
= x −y
with the condition u(x,
1
x
) = 0.
Exercise 11.2
Solve the linear equation
yu
x
+ xu
y
= u
with the Cauchy data u(x, 0) = x
3
.
Exercise 11.3
Solve
x(y
2
+ u)u
x
−y(x
2
+ u)u
y
= (x
2
−y
2
)u
with the Cauchy data u(x, −x) = 1.
Exercise 11.4
Solve
xu
x
+ yu
y
= xe
−u
with the Cauchy data u(x, x
2
) = 0.
Exercise 11.5
Solve the initial value problem
xu
x
+ u
y
= 0, u(x, 0) = f(x)
using the characteristic equations in parametric form.
Exercise 11.6
Solve the initial value problem
u
t
+ au
x
= 0, u(x, 0) = f(x).
Exercise 11.7
Solve the initial value problem
au
x
+ u
y
= u
2
, u(x, 0) = cos x
11 THE CAUCHYPROBLEMFOR FIRST ORDER QUASILINEAR EQUATIONS97
Exercise 11.8
Solve the initial value problem
u
x
+ xu
y
= u, u(1, y) = h(y)
Exercise 11.9
Solve the initial value problem
uu
x
+ u
y
= 0, u(x, 0) = f(x)
Exercise 11.10
Solve the initial value problem

1 −x
2
u
x
+ u
y
= 0, u(x, 0) = sin x
98 FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS
Sample Exam Questions
Exercise 11.11
Consider
xu
x
+ 2yu
y
= 0.
(i) Find and sketch the characteristics.
(ii) Find the solution with u(1, y) = e
y
.
(iii) What happens if you try to ﬁnd the solution satisfying either u(0, y) =
g(y) or u(x, 0) = h(x) for given functions g and h?
(iv) Explain, using your picture of the characteristics, what goes wrong at
(x, y) = (0, 0).
Exercise 11.12
Solve the equation u
x
+ u
y
= u subject to the condition u(x, 0) = cos x.
Exercise 11.13
(a) Find the general solution of the equation
u
x
+ yu
y
= u.
(b) Find the solution satisfying the Cauchy data u(x, 3e
x
) = 2.
(c) Find the solution satisfying the Cauchy data u(x, e
x
) = e
x
.
Exercise 11.14
Solve the Cauchy problem
u
x
+ 4u
y
= x(u + 1)
u(x, 5x) = 1.
Exercise 11.15
Solve the Cauchy problem
u
x
−u
y
= u
u(x, −x) = sin x.
11 THE CAUCHYPROBLEMFOR FIRST ORDER QUASILINEAR EQUATIONS99
Exercise 11.16
(a) Find the characteristics of the equation
yu
x
+ xu
y
= 0.
(b) Sketch some of the characteristics.
(c) Find the solution satisfying the boundary condition u(0, y) = e
−y
2
.
(d) In which region of the plane is the solution uniquely determined?
Exercise 11.17
Consider the equation u
x
+ yu
y
= 0. Is there a solution satisfying the extra
condition
(a) u(x, 0) = 1
(b) u(x, 0) = x?
If yes, give a formula; if no, explain why.
100 FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS
Second Order Linear Partial
Diﬀerential Equations
In this chapter we consider the three fundamental second order linear partial
diﬀerential equations of parabolic, hyperbolic, and elliptic type. These types
arise in many applications such as the wave equation, the heat equation
and the Laplace’s equation. We will study the solvability of each of these
equations.
12 Second Order PDEs in Two Variables
In this section we will brieﬂy review second order partial diﬀerential equa-
tions.
A second order partial diﬀerential equation in the variables x and y
is an equation of the form
F(x, y, u, u
x
, u
y
, u
xx
, u
yy
, u
xy
) = 0. (12.1)
If Equation (12.1) can be written in the form
A(x, y, u, u
x
, u
y
)u
xx
+B(x, y, u, u
x
, u
y
)u
xy
+C(x, y, u, u
x
, u
y
)u
yy
= D(x, y, u, u
x
, u
y
)
(12.2)
then we say that the equation is quasilinear.
If Equation (12.1) can be written in the form
A(x, y)u
xx
+ B(x, y)u
xy
+ C(x, y)u
yy
= D(x, y, u, u
x
, u
y
) (12.3)
then we say that the equation is semilinear.
If Equation (12.1) can be written in the form
A(x, y)u
xx
+B(x, y)u
xy
+C(x, y)u
yy
+D(x, y)u
x
+E(x, y)u
y
+F(x, y)u = G(x, y)
(12.4)
101
102SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
then we say that the equation is linear.
A linear equation is said to be homogeneous when G(x, y) = 0 and non-
homogeneous otherwise.
Equation (12.4) resembles the general equation of a conic section
Ax
2
+ Bxy + Cy
2
+ Dx + Ey + F = 0
which is classiﬁed as either parabolic, hyperbolic, or elliptic based on the sign
of the discriminant B
2
− 4AC. We do the same for a second order linear
partial diﬀerential equation:
• Hyperbolic: This occurs if B
2
−4AC > 0 at a given point in the domain
of u.
• Parabolic: This occurs if B
2
− 4AC = 0 at a given point in the domain
of u.
• Elliptic: This occurs if B
2
− 4AC < 0 at a given point in the domain of
u.
Example 12.1
Determine whether the equation u
xx
+ xu
yy
= 0 is hyperbolic, parabolic or
elliptic.
Solution.
Here we are given A = 1, B = 0, and C = x. Since B
2
− 4AC = −4x, the
given equation is hyperbolic if x < 0, parabolic if x = 0 and elliptic if x > 0
Second order partial diﬀerential equations arise in many areas of scientiﬁc
applications. In what follows we list some of the well-known models that are
of great interest:
1. The heat equation in one-dimensional space is given by
u
t
= ku
xx
where k is a constant.
2. The wave equation in one-dimensional space is given by
u
tt
= c
2
u
xx
where c is a constant.
3. The Laplace equation is given by
∆u = u
xx
+ u
yy
= 0.
12 SECOND ORDER PDES IN TWO VARIABLES 103
Practice Problems
Exercise 12.1
Classify each of the following equation as hyperbolic, parabolic, or elliptic:
(a) Wave propagation: u
tt
= c
2
u
xx
, c > 0
(b) Heat conduction: u
t
= cu
xx
, c > 0.
(c) Laplace’s equation: ∆u = u
xx
+ u
yy
= 0.
Exercise 12.2
Classify the following linear scalar PDE with constant coeﬃcents as hyper-
bolic, parabolic or elliptic.
(a) u
xx
+ 4u
xy
+ 5u
yy
+ u
x
+ 2u
y
= 0
(b) u
xx
−4u
xy
+ 4u
yy
+ 3u
x
+ 4u = 0
(c) u
xx
+ 2u
xy
−3u
yy
+ 2u
x
+ 6u
y
= 0.
Exercise 12.3
Find the region(s) in the xy−plane where the equation
(1 + x)u
xx
+ 2xyu
xy
−y
2
u
yy
= 0
is elliptic, hyperbolic, or parabolic. Sketch these regions.
Exercise 12.4
Show that u(x, t) = cos x sin t is a solution to the problem
u
tt
= u
xx
u(x, 0) = 0
u
t
(x, 0) = cos x
u
x
(0, t) = 0
for all x, t > 0.
Exercise 12.5
Classify each of the following PDE as linear, quasilinear, semi-linear, or non-
linear.
(a) u
t
+ uu
x
= vu
xx
(b) xu
tt
+ tu
yy
+ u
3
u
2
x
= t + 1
(c) u
tt
= c
2
u
xx
(d) u
2
tt
+ u
x
= 0.
104SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
Exercise 12.6
Show that, for all (x, y) = (0, 0), u(x, y) = ln (x
2
+ y
2
) is a solution of
u
xx
+ u
yy
= 0,
and that, for all (x, y, z) = (0, 0, 0), u(x, y, z) =
1

x
2
+y
2
+z
2
is a solution of
u
xx
+ u
yy
+ u
zz
= 0.
Exercise 12.7
Consider the eigenvalue problem
u
xx
= λu, 0 < x < L
u
x
(0) = k
0
u(0)
u
x
(L) = −k
L
u(L)
with Robin boundary conditions, where k
0
and k
L
are given positive numbers
and u = u(x). Can this system have a nontrivial solution u ≡ 0 for λ > 0?
Hint: Multiply the ﬁrst equation by u and integrate over x ∈ [0, L].
Exercise 12.8
Show that u(x, y) = f(x)g(x), where f and g are arbitrary diﬀerentiable
functions, is a solution to the PDE
uu
xy
= u
x
u
y
.
Exercise 12.9
Show that for any n ∈ N, the function u
n
(x, y) = sin nx sinh ny is a solution
to the Laplace equation
∆u = u
xx
+ u
yy
= 0.
Exercise 12.10
Solve
u
xy
= xy.
12 SECOND ORDER PDES IN TWO VARIABLES 105
Sample Exam Questions
Exercise 12.11
Classify each of the following second-oder PDEs according to whether they
are hyperbolic, parabolic, or elliptic:
(a) 2u
xx
−4u
xy
+ 7u
yy
−u = 0
(b) u
xx
−2 cos xu
xy
−sin
2
xu
yy
= 0
(c) yu
xx
+ 2(x −1)u
xy
−(y + 2)u
yy
= 0.
Exercise 12.12
Let c > 0. By computing u
x
, u
xx
, u
t
, and u
tt
show that
u(x, t) =
1
2
(f(x + ct) + f(x −ct)) +
1
2c
_
x+ct
x−ct
g(s)ds
is a solution to the PDE
u
tt
= c
2
u
xx
where f is twice diﬀerentiable function and g is a diﬀerentiable function.
Then compute and simplify u(x, 0) and u
t
(x, 0).
Exercise 12.13
Consider the second-order PDE
yu
xx
+ u
xy
−x
2
u
yy
−u
x
−u = 0.
Determine the region D in R
2
, if such a region exists, that makes this PDE:
(a) hyperbolic, (b) parabolic, (c) elliptic.
Exercise 12.14
Consider the second-order hyperbolic PDE
u
xx
+ 2u
xy
−3u
yy
= 0.
Use the change of variables v(x, y) = y −3x and w(x, y) = x +y to solve the
given equation.
Exercise 12.15
Solve the Cauchy problem
u
xx
+ 2u
xy
−3u
yy
= 0.
u(x, 2x) = 1, u
x
(x, 2x) = x.
106SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
13 Hyperbolic Type: The Wave equation
The wave equation has many physical applications from sound waves in air
to magnetic waves in the Sun’s atmosphere. However, one of the simplest
systems to visualize and describe are waves on a stretched elastic string.
Initially the string is horizontal. Then we distort it by displacing it in the
vertical direction and at some time, say t = 0, we release it and the string
starts to oscillate. The aim is to try and determine the vertical displacement
from the x−axis of the string, u(x, t), as a function of position x and time
t. A displacement of a tiny piece of the string between points P and Q is
shown in Figure 13.1.
Figure 13.1
To derive the wave equation we need to make some simplifying assumptions:
(1) The density of the string, ρ, is constant so that the mass of the string
between P and Q is simply ρ times the length of the string between P and
Q, namely ∆s where
∆s =
_
(∆x)
2
+ (∆u)
2
= ∆x
¸
1 +
_
∆u
∆x
_
2
≈ ∆x
¸
1 +
_
∂u
∂x
_
2
(2) The displacement, u(x, t), and its derivatives are assumed small so that
∆s ≈ ∆x
13 HYPERBOLIC TYPE: THE WAVE EQUATION 107
and the mass of the portion of the string is
ρ∆x.
(3) The only forces acting on this portion of the string are the tensions
T(x, t) at P and T(x + ∆x, t) at Q. (In physics, tension is the magnitude of
the pulling force exerted by a string). The gravitational force is neglected.
(4) The motions are purely vertical. There is no horizontal motion of any
portion of the string.
Next, we consider the forces acting on the typical string portion shown in
Figure 13.1. These forces are:
(i) tension pulling to the right, which has magnitude T(x + ∆x, t), and acts
at an angle θ(x + ∆x, t) above the horizontal.
(ii) tension pulling to the left, which has magnitude T(x, t), and acts at an
angle θ(x, t) above the horizontal.
Now we resolve the forces into their horizontal and vertical components.
• Horizontal: At P the tension force is T(x, t) cos θ(x, t) whereas at Q the
force is T(x + ∆x, t) cos θ(x + ∆x, t). Since there is no horizontal motion,
these forces must balance and so
T(x, t) cos θ(x, t) = T(x + ∆x, t) cos θ(x + ∆x, t) = T.
• Vertical: At P the tension force is −T(x, t) sin θ(x, t) whereas at Q the
force is T(x + ∆x, t) sin θ(x + ∆x, t). Then Newton’s Law of motion gives
mass × acceleration = Applied Forces
that is
ρ∆x

2
u
∂t
2
= T(x + ∆x, t) sin θ(x + ∆x, t) −T(x, t) sin θ(x, t).
Dividing by T we obtain
ρ
T
∆x

2
u
∂t
2
=
T(x + ∆x, t) sin θ(x + ∆x, t)
T(x + ∆x, t) cos θ(x + ∆x, t)

T(x, t) sin θ(x, t)
T(x, t) cos θ(x, t)
=tan θ(x + ∆x, t) −tan θ(x, t).
But
tan θ(x, t) = lim
∆x→0
∆u
∆x
= u
x
(x, t).
108SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
Likewise,
tan θ(x + ∆x, t) = u
x
(x + ∆x, t).
Hence, we get
ρ
T
∆xu
tt
(x, t) = u
x
(x + ∆x, t) −u
x
(x, t).
Dividing by ∆x and letting ∆x →0 we obtain
ρ
T
u
tt
(x, t) = u
xx
(x, t)
or
u
tt
(x, t) = c
2
u
xx
(x, t) (13.1)
where c
2
=
T
ρ
. We call c the wave speed.
D’Alembert Solution of (13.1)
Let v = x +ct and w = x −ct. Then by application of the chain rule we ﬁnd
u
t
=c(u
v
−u
w
)
u
x
=u
v
+ u
w
u
tt
=c
2
(u
vv
−2u
vw
+ u
ww
)
u
xx
=u
vv
+ 2u
vw
+ u
ww
Substituting into (13.1) we obtain
c
2
(u
vv
+ 2u
vw
+ u
ww
) = c
2
(u
vv
−2u
vw
+ u
ww
)
and this simpliﬁes to
4c
2
u
vw
= 0 or u
vw
= 0.
It follows that
u(v, w) = f(v) + g(w)
where f and g are arbitrary diﬀerentiable functions. Now, writing u in terms
of x and y we ﬁnd the general solution
u(x, y) = f(x + ct) + g(x −ct).
D’Alembert’s solution involves two arbitrary functions that are determined
(normally) by two initial conditions.
13 HYPERBOLIC TYPE: THE WAVE EQUATION 109
Example 13.1
Find the solution to the initial value problem
u
tt
=c
2
u
xx
u(x, 0) =v(x)
u
t
(x, 0) =w(x)
Solution.
We have
u(x, 0) = f(x) + g(x) = v(x)
and
u
t
(x, 0) = cf

(x) −cg

(x) = w(x)
which implies that
f(x) −g(x) =
1
c
W(x) =
1
c
_
w(x)dx
Therefore,
g(x) =
1
2
(v(x) −
1
c
W(x))
Hence,
f(x) =
1
2
(v(x) +
1
c
W(x)).
Finally,
u(x, t) =
1
2
[v(x −ct) + v(x + ct) +
1
c
(W(x + ct) −W(x −ct))]
=
1
2
[v(x −ct) + v(x + ct) +
1
c
_
x+ct
x−ct
w(s)ds]
110SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
Practice Problems
Exercise 13.1
Show that if v(x, t) and w(x, t) satisfy equation (13.1) then αv + βw is also
a solution to (13.1), where α and β are constants.
Exercise 13.2
Show that any linear time independent function u(x, t) = ax+b is a solution
to equation (13.1).
Exercise 13.3
Find a solution to (13.1) that satisﬁes the homogeneous conditions u(x, 0) =
u(0, t) = u(L, t) = 0.
Exercise 13.4
Solve the initial value problem
u
tt
=9u
xx
u(x, 0) =cos x
u
t
(x, 0) =0
Exercise 13.5
Solve the initial value problem
u
tt
=u
xx
u(x, 0) =
1
1 + x
2
u
t
(x, 0) =0
Exercise 13.6
Solve the initial value problem
u
tt
=4u
xx
u(x, 0) =1
u
t
(x, 0) =cos (2πx)
13 HYPERBOLIC TYPE: THE WAVE EQUATION 111
Exercise 13.7
Solve the initial value problem
u
tt
=25u
xx
u(x, 0) =v(x)
u
t
(x, 0) =0
where
v(x) =
_
1 if x < 0
0 if x ≥ 0
Exercise 13.8
Solve the initial value problem
u
tt
=c
2
u
xx
u(x, 0) =e
−x
2
u
t
(x, 0) =cos
2
x
Exercise 13.9
Prove that the wave equation, u
tt
= c
2
u
xx
satisﬁes the following properties,
which are known as invariance properties. If u(x, t) is a solution, then
(i) Any translate, u(x −y, t) where y is a ﬁxed constant, is also a solution.
(ii) Any derivative, say u
x
(x, t), is also a solution.
(iii) Any dilation, u(ax, at), is a solution, for any ﬁxed constant a.
Exercise 13.10
Find v(r) if u(r, t) =
v(r)
r
cos nt is a solution to the PDE
u
rr
+
2
r
u
r
= u
tt
.
112SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
Sample Exam Questions
Exercise 13.11
Find the solution of the wave equation on the real line (−∞ < x < +∞)
with the initial conditions
u(x, 0) = e
x
, u
t
(x, 0) = sin x.
Exercise 13.12
The total energy of the string (the sum of the kinetic and potential energies)
is deﬁned as
E(t) =
1
2
_
L
0
(u
2
t
+ c
2
u
2
t
)dx.
(a) Using the wave equation derive the equation of conservation of energy
dE(t)
dt
= c
2
(u
t
(L, t)u
x
(L, t) −u
t
(0, t)u
x
(0, t)).
(b) Assuming ﬁxed ends boundary conditions, that is the ends of the string
are ﬁxed so that u(0, t) = u(L, t) = 0, for all t > 0, show that the energy is
constant.
(c) Assuming free ends boundary conditions for both x = 0 and x = L show
that the energy is constant.
Exercise 13.13
For a wave equation with damping
u
tt
−c
2
u
xx
+ du
t
= 0, d > 0, 0 < x < L
with the ﬁxed ends boundary conditions show that the total energy decreases.
Exercise 13.14
(a) Verify that for any R(x) the function
u(x, t) = R(x −ct)
is a solution of the wave equation u
tt
= c
2
u
xx
. Such solutions are called
traveling waves.
(b) Show that the potential and kinetic energies (see Exercise 13.12) are
equal for the traveling wave solution in (a).
13 HYPERBOLIC TYPE: THE WAVE EQUATION 113
Exercise 13.15
Find the solution of the Cauchy wave equation
u
tt
= 4u
xx
u(x, 0) = x
2
, u
t
(x, 0) = sin 2x.
Simplify your answer as much as possible.
114SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
14 Parabolic Type: The Heat Equation in One-
Dimensional Space
In this section, We will look at a model for describing the distribution of
temperature in a solid material as a function of time and space.
Before we begin our discussion of the mathematics of the heat equation, we
must ﬁrst determine what is meant by the term heat? Heat is type of energy
known as thermal energy. Heat travels in waves like other forms of energy,
and can change the matter it touches. It can heat it up and cause chemical
reactions like burning to occur.
Heat can be released through a chemical reaction (such as the nuclear re-
actions that make the Sun “burn”) or can be trapped for a limited time by
insulators. It is often released along with other kinds of energy such as light
waves or sound waves. For example, a burning candle releases light and heat
waves. On the other hand, an explosion releases light, heat, and sound waves.
The most common units of heat are BTU (British Thermal Unit), Calorie
and Joule.
Consider now a rod made of homogeneous heat conducting material of uni-
form density ρ and constant cross section A, placed along the x−axis from
x = 0 to x = L as shown in Figure 14.1.
Figure 14.1
Assume the heat ﬂows only in the x−direction, with the lateral sides insu-
lated, and the only way heat can enter or leave the rod is at either end. Also
we assume that the temperature of the rod is constant at any point of the
cross section. In other words, temperature will only vary in x and we can
hence consider the rod to be a one spatial dimensional rod. We will also
assume that heat energy in any piece of the rod is conserved.
Let u(x, t) be the temperature of the cross section at the point x and the
time t. Consider an inﬁnitesimal portion U of the rod from x to x + ∆x of
length ∆x as shown in Figure 14.2.
14 PARABOLIC TYPE: THE HEAT EQUATIONINONE-DIMENSIONAL SPACE115
Figure 14.2
From the theory of heat conduction, the rate of change of heat in the portion
U is given by
_
x+∆x
x
cρAu
t
(s, t)dx
where c is the speciﬁc heat, that is, the amount of heat energy that it takes
to raise one unit of mass of the material by one unit of temperature.
Assuming that u is continuously diﬀerentiable, we can apply the mean value
theorem for integrals and ﬁnd x ≤ ξ ≤ x + ∆x such that
_
x+∆x
x
u
t
(s, t)dx = ∆xu
t
(ξ, t).
Thus, the rate of change of heat in U is given by
cρA∆xu
t
(ξ, t).
On the other hand, by Fourier (or Fick’s) law of heat conduction, the rate
of heat ﬂow through any cross section is proportional to the area A and the
gradient of the temperature normal to the cross section, and heat ﬂows in
the direction of decreasing temperature. Thus, the rate of heat ﬂowing in U
through the cross section at x is −KAu
x
(x, t) and the rate of heat ﬂowing
out of U through the cross section at x+∆x is −KAu
x
(x+∆x, t), where K
is the thermal conductivity of the rod.
Now, the conservation of energy law states
rate of change of heat in U = rate of heat ﬂowing in − rate of heat ﬂowing
out
116SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
or mathematically written as,
cρA∆xu
t
(ξ, t) = −KAu
x
(x, t) + KAu
x
(x + ∆x, t)
or
cρA∆xu
t
(ξ, t) = KA[u
x
(x + ∆x, t) −u
x
(x, t)].
Dividing this last equation by cAρ∆x and letting ∆x →0 we obtain
u
t
(x, t) = ku
xx
(x, t) (14.1)
where k =
K

is called the diﬀusivity constant.
Equation (14.1) is the one dimensional heat equation which is second order,
linear, homogeneous, and of parabolic type.
The non-homogeneous heat equation
u
t
= ku
xx
+ f(x)
is known as the heat equation with an external heat source f(x). An ex-
ample of an exterenal heat source is the heat generated from a candle placed
under the bar.
The function
E(t) =
_
L
0
cρu(x, t)dx
is called the total thermal energy at time t of the entire rod.
Example 14.1
The two ends of a uniform rod of length L are insulated. There is a con-
stant source of thermal energy q
0
= 0 and the temperature is initially
u(x, 0) = f(x).
(a) Write the equation and the boundary conditions for this model.
(b) Calculate the total thermal energy of the entire rod.
Solution.
(a) The model is given by the PDE
cρu
t
(x, t) = Ku
xx
+ q
0
with boundary conditions
u
x
(0, t) = u
x
(L, t) = 0.
14 PARABOLIC TYPE: THE HEAT EQUATIONINONE-DIMENSIONAL SPACE117
(b) First note that
d
dt
_
L
0
cρu(x, t)dx =
_
L
0
cρu
t
(x, t)dx =
_
L
0
Ku
xx
dx +
_
L
0
q
0
dx
= Ku
x
|
L
0
+ q
0
L = 0
since u
x
(0, t) = u
x
(L, t) = 0. Integrating in time from 0 to t we ﬁnd
E(t) = q
0
Lt + C.
But C = E(0) =
_
L
0
cρu(x, 0)dx =
_
L
0
cρf(x)dx. Hence, the total thermal
energy is given by
E(t) =
_
L
0
cρf(x)dx + q
0
Lt
Initial Boundary Value Problems
In order to solve the heat equation we must give the problem some initial
conditions. If you recall from the theory of ODE, the number of conditions
required for solving initial value problems always matched the highest order
of the derivative in the equation.
In partial diﬀerential equations the same idea holds except now we have to
pay attention to the variable we are diﬀerentiating with respect to as well.
So, for the heat equation we have got a ﬁrst order time derivative and so we
will need one initial condition and a second order spatial derivative and so
we will need two boundary conditions.
For the initial condition, we deﬁne the temperature of every point along the
rod at time t = 0 by
u(x, 0) = f(x)
where f is a given (prescribed) function of x. This function is known as the
initial temperature distribution.
The boundary conditions will tell us something about what the temperature
is doing at the ends of the bar. The conditions are given by
u(0, t) = T
0
and u(L, t) = T
L
.
and they are called as the Dirichlet conditions. In this case, the general
form of the heat equation initial boundary value problem is to ﬁnd u(x, t)
satisfying
u
t
(x, t) =ku
xx
(x, t), 0 ≤ x ≤ L, t > 0
u(x, 0) =f(x), 0 ≤ x ≤ L
u(0, t) =T
0
, u(L, t) = T
L
, t > 0.
118SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
In the case of insulated endpoints, i.e. there is no heat ﬂow out of them, we
use the boundary conditions
u
x
(0, t) = u
x
(L, t) = 0.
These conditions are examples of what is known as Neumann boundary
conditions. In this case, the general form of the heat equation initial bound-
ary value problem is to ﬁnd u(x, t) satisfying
u
t
(x, t) =ku
xx
(x, t), 0 ≤ x ≤ L, t > 0
u(x, 0) =f(x), 0 ≤ x ≤ L
u
x
(0, t) =u
x
(L, t) = 0, t > 0.
14 PARABOLIC TYPE: THE HEAT EQUATIONINONE-DIMENSIONAL SPACE119
Practice Problems
Exercise 14.1
Show that if u(x, t) and u(x, t) satisfy equation (14.1) then αu +βv is also a
solution to (14.1), where α and β are constants.
Exercise 14.2
Show that any linear time independent function u(x, t) = ax+b is a solution
to equation (14.1).
Exercise 14.3
Find a lnear time independent solution u to (14.1) that satisﬁes u(0, t) = T
0
and u(L, T) = T
L
.
Exercise 14.4
Show that to solve (14.1) with the boundary conditions u(0, t) = T
0
and
u(L, t) = T
L
it suﬃces to solve (14.1) with the homogeneous boundary
conditions u(0, t) = u(L, t) = 0.
Exercise 14.5
Find a solution to (14.1) that satisﬁes the conditions u(x, 0) = u(0, t) =
u(L, t) = 0.
Exercise 14.6
Let (I) denote equation (14.1) together with intial time condition u(x, 0) =
f(x), where f is not the zero function, and the homogeneous boundary condi-
tions u(0, t) = u(L, t) = 0. Suppose a nontrivial solution to (I) can be written
in the form u(x, t) = X(x)T(t). Show that X and T satisﬁes the ODE
X

λ
k
X = 0 and T

−λT = 0.
for some constant λ.
Exercise 14.7
Consider again the solution u(x, t) = X(x)T(t). Clearly, T(t) = T(0)e
−λt
.
Suppose that λ > 0.
(a) Show that X(x) = Ae

αx
+ Be

αx
, where α =
λ
k
and A and B are
arbitrary constants.
(b) Show that A and B satisﬁes the two equations A+B = 0 and A(e

λL

e

λL
) = 0.
120SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
(c) Show that A = 0 leads to a contradiction.
(d) Using (b) and (c) show that e

λL
= e

λL
. Show that this equality leads
to a contradiction. We conclude that λ < 0.
Exercise 14.8
Consider the results of the previous exercise.
(a) Show that X(x) = c
1
cos βx + c
2
sin βx where β =
_
−λ
k
.
(b) Show that λ = λ
n
= −
kn
2
π
2
L
2
, where n is an integer.
Exercise 14.9
Show that u(x, t) =

n
k=1
u
k
(x, t), where u
n
(x, t) = c
n
e
kn
2
π
2
L
2
t
sin
_

L
_
x sat-
isﬁes (14.1) and the homogeneous boundary conditions.
Exercise 14.10
Suppose that a wire is stretched between 0 and a. Describe the boundary
conditions for the temperature u(x, t) when
(i) the left end is kept at 0 degrees and the right end is kept at 100 degrees;
and
(ii) when both ends are insulated.
Exercise 14.11
Let u
t
= u
xx
for 0 < x < π and t > 0 with boundary conditions u(0, t) =
0 = u(π, t) and initial condition u(x, 0) = f(x). Let E(t) =
_
π
0
(u
2
t
+ u
2
x
)dx.
Show that E

(t) < 0.
Exercise 14.12
Suppose
u
t
= u
xx
+ 4, u
x
(0, t) = 5, u
x
(L, t) = 6, u(x, 0) = f(x).
Calculate the total thermal energy of the one-dimensional rod (as a function
of time).
14 PARABOLIC TYPE: THE HEAT EQUATIONINONE-DIMENSIONAL SPACE121
Sample Exam Questions
Exercise 14.13
Consider the heat equation
u
t
= ku
xx
for x ∈ (0, 1) and t > 0, with boundary conditions u(0, t) = 2 and u(1, t) = 3
for t > 0 and initial conditions u(x, 0) = x for x ∈ (0, 1). A function v(x)
that satisﬁes the equation v

(x) = 0, with conditions v(0) = 2 and v(1) = 3
is called a steady-state solution. That is, the steady-state solutions of the
heat equation are those solutions that don’t depend on time. Find v(x).
Exercise 14.14
Consider the equation for the one-dimensional rod of length L with given
heat energy source:
u
t
= u
xx
+ q(x).
Assume that the initial temperature distribution is given by u(x, 0) = f(x).
Find the equilibrium (steady state) temperature distribution in the following
cases.
(a) q(x) = 0, u(0) = 0, u(L) = T.
(b) q(x) = 0, u
x
(0) = 0, u(L) = T.
(c) q(x) = 0, u(0) = T, u
x
(L) = α.
(d) q(x) = 1, u(0) = T
1
, u(L) = T
2
.
Exercise 14.15
Two rods of diﬀerent materials (thermal diﬀusion constants k
1
and k
2
; as-
sume cρ = 1 in both cases) are joined at x = L and are in perfect thermal
contact (this means that the temperature is continuous at x = L and the
heat energy ﬂowing out of one end ﬂows into the other). (See ﬁgure below.)
(a) Assume that the temperature distribution is steady (u = u(x), does
not depend on time), and satisﬁes u(0) = T
1
, u(2L) = T
2
. Find u(x).
(b) Let L = 1, T
1
= 0, T
2
= 100, k
1
= 2, k
2
= 1. Plot the temperature u as a
function of x.
122SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
Exercise 14.16
Consider the equation for the one-dimensional rod of length L with insulated
ends:
cρu
t
= Ku
xx
, u
x
(0, t) = u
x
(L, t) = 0.
(a) Give the expression for the total thermal energy of the rod.
(b) Show using the equation and the boundary conditions that the total
thermal energy is constant.
Exercise 14.17
Suppose
u
t
= u
xx
+ x, u(x, 0) = f(x), u
x
(0, t) = β, u
x
(L, t) = 7.
(a) Calculate the total thermal energy of the one-dimensional rod (as a func-
tion of time).
(b) From part (a) ﬁnd the value of β for which a steady-state solution exist.
(c) For the above value of β ﬁnd the steady state solution.
15 AN INTRODUCTION TO FOURIER SERIES 123
15 An Introduction to Fourier Series
In this and the next section we will have a brief look to the subject of Fourier
series. The point here is to do just enough to allow us to do some basic so-
lutions to partial diﬀerential equations later in the book.
Motivation: In Calculus we have seen that certain functions may be repre-
sented as power series by means of the Taylor expansions. These functions
must have inﬁnitely many derivatives, and the series provide a good approx-
imation only in some (often small) vicinity of a reference point.
Fourier series constructed of trigonometric rather than power functions, and
can be used for functions not only not diﬀerentiable, but even discontinuous
at some points. The main limitation of Fourier series is that the underlying
function should be periodic.
Recall from calculus that a function series is a series where the summands
are functions. Examples of function series include power series, Laurent se-
ries, Fourier series, etc.
Unlike series of numbers, there exist many types of convergence of series of
functions, namely, pointwise, uniform, etc. We say that a series of functions

n=1
f
n
(x) converges pointwise to a function f if and only if the sequence
of partial sums
S
n
(x) = f
1
(x) + f
2
(x) +· · · + f
n
(x)
converges pointwise to f. We write

n=1
f
n
(x) = lim
n→∞
S
n
(x) = f(x).
Likewise, we say that a series of functions

n=1
f
n
(x) converges uniformly
to a function f if and only if the sequence of partial sums {S
n
}

n=1
converges
uniformly to f.
In this section we introduce a type of series of functions known as Fourier
series. They are given by
f(x) =
a
0
2
+

n=1
_
a
n
cos
_

L
x
_
+ b
n
sin
_

L
x
__
, −L ≤ x ≤ L (15.1)
where a
n
and b
n
are called the Fourier coeﬃcients. The expression on the
right is called a trigonometric series. Note that we begin the series with
a
0
2
as opposed to simply a
0
to simplify the coeﬃcient formula for a
n
that we
124SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
will derive later in this section.
The main questions we want to consider next are the questions of determin-
ing which functions can be represented by Fourier series and if so how to
compute the coeﬃcients a
n
and b
n
.
Before answering these questions, we look at some of the properties of Fourier
series.
Periodicity Property
Recall that a function f is said to be periodic with period T > 0 if
f(x + T) = f(x) for all x, x + T in the domain of f. The smallest value
of T for which f is periodic is called the fundamental period. A graph of
a T−periodic function is shown in Figure 15.1.
Figure 15.1
For a T−periodic function we have
f(x) = f(x + T) = f(x + 2T) = · · · .
Note that the deﬁnite integral of a T−periodic function is the same over any
interval of length T. By Exercise 15.1 below, if f and g are two periodic func-
tions with common period T, then the product fg and an arbitrary linear
combination c
1
f +c
2
g are also periodic with period T. It is an easy exercise
to show that the Fourier series (15.1) is periodic with fundamental period 2L.
Orthogonality Property
Recall from Calculus that for each pair of vectors u and v we associate a
scalar quantity u· v called the dot product of u and v. We say that u and v
are orthogonal if and only if u · v = 0. We want to deﬁne a similar concept
for functions.
Let f and g be two functions with domain the closed interval [a, b]. We deﬁne
15 AN INTRODUCTION TO FOURIER SERIES 125
a function that takes a pair of functions to a scalar. Symbolically, we write
< f, g >=
_
b
a
f(x)g(x)dx.
We call < f, g > the inner product of f and g. We say that f and g
are orthogonal if and only if < f, g >= 0. A set of functions is said to
be mutually orthogonal if each distinct pair of functions in the set is
orthogonal.
Example 15.1
Show that the set
_
1, cos
_

L
x
_
, sin
_

L
x
_
: n ∈ N
_
is mutually orthogonal in
[−L, L].
Solution.
We have
_
L
−L
1 · cos
_

L
x
_
dx =
_
sin
_

L
x
__
L
−L
= 0
and
_
L
−L
1 · sin
_

L
x
_
dx = −
_
cos
_

L
x
__
L
−L
= 0.
Now, for n = m we have
_
L
−L
cos
_

L
x
_
cos
_

L
x
_
dx =
1
2
_
L
−L
_
cos
_
(m + n)π
L
x
_
+ cos
_
(m−n)π
L
x
__
dx
=
1
2
_
L
(m + n)π
sin
_
(m + n)π
L
x
_
+
L
(m−n)π
sin
_
(m−n)π
L
x
__
L
−L
= 0
where we used the trigonometric identity
cos a cos b =
1
2
[cos (a + b) + cos (a −b)].
In the exercises below, we show that
_
L
−L
sin
_

L
x
_
sin
_

L
x
_
dx = 0
126SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
and
_
L
−L
cos
_

L
x
_
sin
_

L
x
_
dx = 0
The reason we care about these functions being orthogonal is because we will
exploit this fact to develop a formula for the coeﬃcients in our Fourier series.
Now, in order to answer the ﬁrst question mentioned earlier, that is, which
functions can be expressed as a Fourier series expansion, we need to intro-
duce some mathematical concepts.
A function f(x) is said to be piecewise continuous on [a, b] if it is contin-
uous in [a, b] execept possibly at ﬁnitely many points of discontinuity within
the interval [a, b], and at each point of discontinuity, the right- and left-
handed limits of f exist. An example of a piecewise continuous function is
the function
f(x) =
_
x 0 ≤ x < 1
x
2
−x 1 ≤ x ≤ 2
We will say that f is piecewise smooth in [a, b] if and only if f(x) as well
as its derivatives are piecewise continuous.
The following theorem, proven in more advanced books, ensures that a
Fourier decomposition can be found for any function which is piecewise
smooth.
Theorem 15.1
Let f be a 2L-periodic function. If f is a piecewise smooth on [−L, L] then
for all points of discontinuity x ∈ (−L, L) we have
f(x

) + f(x
+
)
2
=
a
0
2
+

n=1
_
a
n
cos
_

L
x
_
+ b
n
sin
_

L
x
__
.
where as for points of continuity x ∈ (−L, L) we have
f(x) =
a
0
2
+

n=1
_
a
n
cos
_

L
x
_
+ b
n
sin
_

L
x
__
.
Remark 15.1
(1) Almost all functions occurring in practice are piecewise smooth functions.
(2) Given a non-periodic function f on [−L, L]. The above theorem applies
to the periodic extension F of f where F(x + 2nL) = f(x) (n ∈ Z) and
F(x) = f(x) on [−L, L].
15 AN INTRODUCTION TO FOURIER SERIES 127
Convergence Results of Fourier Series
We list few of the results regarding the convergence of Fourier series:
(1) The type of convergence in the above theorem is pointwise convergence.
(2) The convergence is uniform for a continuous function f on [−L, L] such
that f(−L) = f(L).
(3) The convergence is uniform whenever

n=1
(|a
n
|
2
+|b
n
|
2
) is convergent.
(4) If f(x) is periodic, continuous, and has a piecewise continuous derivative,
then the Fourier Series corresponding to f converges uniformly to f(x) for
the entire real line.
(5) The convergence is uniform on any closed interval that does not contain
a point of discontinuity.
Euler-Fourier Formulas
Next, we will answer the second question mentioned earlier, that is, the ques-
tion of ﬁnding formulas for the coeﬃcients a
n
and b
n
. These formulas for a
n
and b
n
are called Euler-Fourier formulas which we derive next. We will as-
sume that the RHS in (15.1) converges uniformly to f(x) on the interval
[−L, L]. Integrating both sides of (15.1) we obtain
_
L
−L
f(x)dx =
_
L
−L
a
0
2
dx +
_
L
−L

n=1
_
a
n
cos
_

L
x
_
+ b
n
sin
_

L
x
__
dx.
Since the trigonometric series is assumed to be uniformly convergent, from
the previous section, we can interchange the order of integration and sum-
mation to obtain
_
L
−L
f(x)dx =
_
L
−L
a
0
2
dx +

n=1
_
L
−L
_
a
n
cos
_

L
x
_
+ b
n
sin
_

L
x
__
dx.
But
_
L
−L
cos
_

L
x
_
dx =
L

sin
_

L
x
__
L
−L
= 0
and likewise
_
L
−L
sin
_

L
x
_
dx = −
L

cos
_

L
x
__
L
−L
= 0.
Thus,
a
0
=
1
L
_
L
−L
f(x)dx.
128SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
To ﬁnd the other Fourier coeﬃcients, we recall the results of Exercises 15.2
- 15.3 below.
_
L
−L
cos
_

L
x
_
cos
_

L
x
_
dx =
_
L if m = n
0 if m = n
_
L
−L
sin
_

L
x
_
sin
_

L
x
_
dx =
_
L if m = n
0 if m = n
_
L
−L
sin
_

L
x
_
sin
_

L
x
_
dx = 0, ∀m, n.
Now, to ﬁnd the formula for the Fourier coeﬃcients a
m
for m > 0, we multiply
both sides of (15.1) by cos
_

L
x
_
and integrate from −L to L to otbain
_
L
−L
f(x) cos
_

L
x
_
=
_
L
−L
a
0
2
cos
_

L
x
_
dx +

n=1
_
a
n
_
L
−L
cos
_

L
x
_
cos
_

L
x
_
dx
+ b
n
_
L
−L
sin
_

L
x
_
cos
_

L
x
_
_
dx.
Hence,
_
L
−L
f(x) cos
_

L
x
_
dx = a
m
L
and therefore
a
m
=
1
L
_
L
−L
f(x) cos
_

L
x
_
dx.
Likewise, we can show that
b
m
=
1
L
_
L
−L
f(x) sin
_

L
x
_
dx.
Example 15.2
Find the Fourier series expansion of
f(x) =
_
0, x ≤ 0
x, x > 0.
on the interval [−π, π].
15 AN INTRODUCTION TO FOURIER SERIES 129
Solution.
We have
a
0
=
1
π
_
π
−π
f(x)dx =
1
π
_
π
0
xdx =
π
2
a
n
=
1
π
_
π
0
x cos nxdx =
1
π
_
x sin nx
n
+
cos nx
n
2
_
π
0
=
(−1)
n
−1
πn
2
b
n
=
1
π
_
π
0
x sin nxdx =
1
π
_

x cos nx
n
+
sin nx
n
2
_
π
0
=
(−1)
n+1
n
Hence,
f(x) =
π
4
+

n=1
_
(−1)
n
−1
πn
2
cos (nx) +
(−1)
n+1
n
sin (nx)
_
Example 15.3
Apply the Theorem 15.1 to the function in Example 15.2.
Solution.
Let F be a periodic extension of f of period 2π. Thus, f(x) = F(x) on the
interval [−π, π]. Clearly, F is a piecewise smooth function so that by the
previous thereom we can write
π
4
+

n=1
_
(−1)
n
−1
πn
2
cos (nx) +
(−1)
n+1
n
sin (nx)
_
=
_
_
_
π
2
, if x = −π
f(x), if −π < x < π
π
2
, if x = π
Taking x = π we have the identity
π
4
+

n=1
(−1)
n
−1
πn
2
(−1)
n
=
π
2
which can be simpliﬁed to

n=1
1
(2n −1)
2
=
π
2
8
.
This provides a method for computing an approximate value of π
130SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
Remark 15.2
An example of a function that does not a Fourier series representation is
the function f(x) =
1
x
2
on [−L, L]. For example, the coeﬃcient a
0
for this
function does not exist. Thus, not every function can be written as a Fourier
series expansion.
The ﬁnal topic of discussion here is the topic of diﬀerentiation and integration
of Fourier series. In particular we want to know if we can diﬀerentiate a
Fourier series term by term and have the result be the Fourier series of the
derivative of the function. Likewise we want to know if we can integrate a
Fourier series term by term and arrive at the Fourier series of the integral of
the function. Answers to these questions are provided next.
Theorem 15.2
A Fourier series of a piecewise smooth function f can always be integrated
term by term and the result is a convergent inﬁnite series that always con-
verges to
_
L
−L
f(x)dx even if the original series has jumps.
Theorem 15.3
A Fourier series that is continuous can be diﬀerentiated term by term if f

(x)
is piecewise smooth. The result of the diﬀerentiation is the Fourier series of
f

(x).
15 AN INTRODUCTION TO FOURIER SERIES 131
Practice Problems
Exercise 15.1
Let f and g be two functions with common domain D and common period
T. Show that
(a) fg is periodic of period T.
(b) c
1
f + c
2
g is periodic of period T, where c
1
and c
2
are real numbers.
Exercise 15.2
Show that for m = n we have
(a)
_
L
−L
sin
_

L
x
_
sin
_

L
x
_
dx = 0 and
(b)
_
L
−L
cos
_

L
x
_
sin
_

L
x
_
dx = 0
Exercise 15.3
Compute the following integrals:
(a)
_
L
−L
cos
2
_

L
x
_
dx.
(b)
_
L
−L
sin
2
_

L
x
_
dx.
(c)
_
L
−L
cos
_

L
x
_
sin
_

L
x
_
dx.
Exercise 15.4
Find the Fourier coeﬃcients of
f(x) =
_
_
_
−π, −π ≤ x < 0
π, 0 < x < π
0, x = 0, π
on the interval [−π, π].
Exercise 15.5
Find the Fourier series of f(x) = x
2

1
2
on the interval [−1, 1].
Exercise 15.6
Find the Fourier series of the function
f(x) =
_
_
_
−1, −2π < x < −π
0, −π < x < π
1, π < x < 2π.
132SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
Exercise 15.7
Find the Fourier series of the function
f(x) =
_
1 + x, −2 ≤ x ≤ 0
1 −x, 0 < x ≤ 2.
Exercise 15.8
Show that f(x) =
1
x
is not piecewise continuous on [−1, 1].
Exercise 15.9
Assume that f(x) is continuous and has period 2L. Prove that
_
L
−L
f(x)dx =
_
L+a
−L+a
f(x)dx
is independent of a ∈ R. In particular, it does not matter over which interval
the Fourier coeﬃcients are computed as long as the interval length is 2L.
[Remark: This result is also true for piecewise continuous functions].
Exercise 15.10
Consider the function f(x) deﬁned by
f(x) =
_
1 0 ≤ x < 1
2 1 ≤ x < 3
and extended periodically with period 3 to R so that f(x +3) = f(x) for all
x.
(i) Find the Fourier series of f(x).
(ii) Discuss its limit: In particular, does the Fourier series converge pointwise
or uniformly to its limit, and what is this limit?
(iii) Plot the graph of f(x) and the limit of the Fourier series.
15 AN INTRODUCTION TO FOURIER SERIES 133
Sample Exam Questions
Exercise 15.11
For the following functions f(x) on the interval −L < x < L, determine the
coeﬃcients a
n
, n = 0, 1, 2, · · · and b
n
, n ∈ N of the Fourier series expansion.
(a) f(x) = 1.
(b) f(x) = 2 + sin
_
pix
L
_
.
(c) f(x) =
_
1 x ≤ 0
0 x > 0
(d) f(x) = x.
Exercise 15.12
Let f(t) be the function with period 2π deﬁned as
f(t) =
_
2 if 0 ≤ x ≤
π
2
0 if
π
2
< x ≤ 2π
f(t) has a Fourier series and that series is equal to
a
0
2
+

n=1
∞(a
n
cos nt + b
n
sin nt).
Find
a
3
b
3
.
Exercise 15.13
Let f(x) = x
3
on [−π, π], extended periodically to all of R. Find the Fourier
coeﬃcients a
n
, n = 1, 2, 3, · · · .
Exercise 15.14
Let f(x) be the square wave function
f(x) =
_
−π −π ≤ x < 0
π 0 ≤ x ≤ π
extended periodically to all of R. To what value does the Fourier series of
f(x) converge when x = 0?
Exercise 15.15
(a) Find the Fourier series of
f(x) =
_
1 −π ≤ x < 0
2 0 ≤ x ≤ π
extended periodically to all of R. Simplify your coeﬃcients as much as pos-
sible.
(b) Use (a) to evaluate the series

n=1
(−1)
n+1
(2n−1)
. Hint: Evaluate the Fourier
series at x =
π
2
.
134SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
16 Fourier Sines Series and Fourier Cosines Se-
ries
In this section we discuss some important properties of Fourier series when
the underlying function f is either even or odd.
A function f is odd if it satisﬁes f(−x) = f(x) for all x in the domain of f
whereas f is even if it satisﬁes f(−x) = f(x) for all x in the domain of f.
Now, we recall from Exercises (1.6)-(1.7) the following facts about even and
odd functions. If f(x) is even then
_
L
−L
f(x)dx = 2
_
L
0
f(x)dx.
If f is odd then
_
L
−L
f(x)dx = 0.
Using just these basic facts we can ﬁgure out some important properties of
the Fourier series we get for odd or even functions.
Example 16.1
Show the following
(a) If f and g are either both even or both odd then fg is even.
(b) If f is odd and g is even then fg is odd.
Solution.
(a) Suppose that both f and g are even. Then (fg)(−x) = f(−x)g(−x) =
f(x)g(x) = (fg)(x). That is, fg is even. Now, suppose that both f and g
are odd. Then (fg)(−x) = f(−x)g(−x) = [−f(x)][−g(x)] = (fg)(x). That
is, fg is even.
(b) f is odd and g is even. Then (fg)(−x) = f(−x)g(−x) = −f(x)g(x) =
−(fg)(x). That is, fg is odd
Example 16.2
(a) Find the value of the integral
_
L
−L
f(x) sin
_

L
x
_
dx when f is even.
(b) Find the value of the integral
_
L
−L
f(x) cos
_

L
x
_
dx when f is odd.
Solution.
(a) Since the function sin
_

L
x
_
is odd and f is even, we have that f(x) sin
_

L
x
_
16 FOURIER SINES SERIES AND FOURIER COSINES SERIES 135
is odd so that
_
L
−L
f(x) sin
_

L
x
_
dx = 0
(b) Since the function cos
_

L
x
_
is even and f is odd, we have that f(x) cos
_

L
x
_
is odd so that
_
L
−L
f(x) cos
_

L
x
_
dx = 0
Even and Odd Extensions
Let f : [0, L] → R be a piecewise smooth function. We deﬁne the odd
extension of this function on the interval −L ≤ x ≤ L by
f
odd
(x) =
_
_
_
f(x) 0 < x ≤ L
−f(−x) −L ≤ x < 0.
0, x = 0
This function will be odd on the interval [−L, L], and will be equal to f(x)
on the interval [0, L]. We can then further extend this function to the entire
real line by deﬁning it to be 2L periodic. Let f
odd
denote this extension. We
note that f
odd
is an odd function and piecewise smooth so that by Theorem
15.1 it possesses a Fourier series expansion, and from the fact that it is odd
all of the a

n
s are zero. Moreover, in the interval [0, L] we have
f(x) =

n=1
b
n
sin
_

L
x
_
. (16.1)
We call (16.1) the Fourier sine series of f.
The coeﬃcients b
n
are given by the formula
b
n
=
1
L
_
L
−L
f
odd
sin
_

L
x
_
dx =
2
L
_
L
0
f
odd
sin
_

L
x
_
dx
=
2
L
_
L
0
f(x) sin
_

L
x
_
dx
since f
odd
sin
_

L
x
_
is an even function.
Likewise, we can deﬁne the even extension of f on the interval −L ≤ x ≤ L
by
f
even
(x) =
_
f(x) 0 ≤ x ≤ L
f(−x) −L ≤ x ≤ 0
136SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
We can then further extend this function to the entire real line by deﬁning
it to be 2L periodic. Let f
even
denote this extension. Again, we note that
f
even
is equal to the original function f(x) on the interval upon which f(x)
is deﬁned. Since f
even
is piecewise smooth, by Theorem 15.1 it possesses a
Fourier series expansion, and from the fact that it is even all of the b

n
s are
zero. Moreover, in the interval [0, L] we have
f(x) =
a
0
2
+

n=1
a
n
cos
_

L
x
_
. (16.2)
We call (16.2) the Fourier cosine series of f. The coeﬃcients a
n
are given
by
a
n
=
2
L
_
L
0
f(x) cos
_

L
x
_
dx.
Example 16.3
Graph the odd and even extensions of the function f(x) = x, 0 ≤ x ≤ 1.
Solution.
We have f
odd
(x) = x for −1 ≤ x ≤ 1. The odd extension of f is shown in
Figure 18.1(a). Likewise,
f
even
(x) =
_
x 0 ≤ x ≤ 1
−x −1 ≤ x ≤ 0
The even extension is shown in Figure 16.1(b)
Figure 16.1
16 FOURIER SINES SERIES AND FOURIER COSINES SERIES 137
Example 16.4
Find the Fourier sine series of the function
f(x) =
_
x, 0 ≤ x ≤
π
2
π −x,
π
2
≤ x ≤ π
Solution.
We have
b
n
=
2
π
_
_ π
2
0
x sin nxdx +
_
π
π
2
(π −x) sin nxdx
_
.
Using integration by parts we ﬁnd
_ π
2
0
x sin nxdx =
_

x
n
cos nx

2
0
+
1
n
_ π
2
0
cos nxdx
=−
π cos (nπ/2)
2n
+
1
n
2
[sin nx]
π
2
0
=−
π cos (nπ/2)
2n
+
sin (nπ/2)
n
2
while
_
π
π
2
(π −x) sin nxdx =
_

(π −x)
n
cos nx
_
π
π
2

1
n
_
π
π
2
cos nxdx
=
π cos (nπ/2)
2n

1
n
2
[sin nx]
π
π
2
=
π cos (nπ/2)
2n
+
sin (nπ/2)
n
2
Thus,
b
n
=
4 sin (nπ/2)
πn
2
,
and the Fourier sine series of f(x) is
f(x) =

n=1
4 sin (nπ/2)
πn
2
sin nx =

n=1
4(−1)
2n−1
π(2n −1)
2
sin (2n −1)x
138SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
Practice Problems
Exercise 16.1
Give an example of a function that is both even and odd.
Exercise 16.2
Graph the odd and even extensions of the function f(x) = 1, 0 ≤ x ≤ 1.
Exercise 16.3
Graph the odd and even extensions of the function f(x) = L−x for 0 ≤ x ≤
L.
Exercise 16.4
Graph the odd and even extensions of the function f(x) = 1 + x
2
for 0 ≤
x ≤ L.
Exercise 16.5
Find the Fourier cosine series of the function
f(x) =
_
x, 0 ≤ x ≤
π
2
π −x,
π
2
≤ x ≤ π
Exercise 16.6
Find the Fourier cosine series of f(x) = x on the interval [0, π].
Exercise 16.7
Find the Fourier sine series of f(x) = 1 on the interval [0, π].
Exercise 16.8
Find the Fourier sine series of f(x) = cos x on the interval [0, π].
Exercise 16.9
Find the Fourier cosine series of f(x) = e
2x
on the interval [0, 1].
Exercise 16.10
Consider the function f(x) = x on the interval [0, 2].
(i) Write down the deﬁnition of the Fourier sine and cosine series for f,
including formulas for the coeﬃcients in terms of f. (Note: you do not need
to compute the coeﬃcients by carrying out the integration. Just write down
their formulas in terms of f.)
16 FOURIER SINES SERIES AND FOURIER COSINES SERIES 139
(ii) What is the limit of the Fourier sine series of f on (0, 2) and in what
sense does it converge to its limit?
(iii) What is the limit of the Fourier cosine series of f on (0, 2) and in what
sense does it converge to its limit?
(iv) Plot the even extension of f(x), the odd extension of f(x), the limit of
its Fourier sine series, and the limit of its Fourier cosine series on the interval
[−2, 2]. (Don’t forget to consider the endpoints 2 and −2.)
140SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
Sample Exam Questions
Exercise 16.11
For the following functions on the interval 0 < x < L, ﬁnd the coeﬃcients b
n
of the Fourier sine expansion.
(a) f(x) = sin
_

L
x
_
.
(b) f(x) = 1
(c) f(x) = cos
_
π
L
x
_
.
Exercise 16.12
For the following functions on the interval −L < x < L, ﬁnd the coeﬃcients
a
n
of the Fourier cosine expansion.
(a) f(x) = 5 + cos
_

L
x
_
.
(b) f(x) = x
(c)
f(x) =
_
1 0 < x ≤
L
2
0 x >
L
2
Exercise 16.13
Consider a function f(x), deﬁned on 0 ≤ x ≤ L, which is even (symmetric)
around x =
L
2
. Show that the even coeﬃcients (n even) of the Fourier sine
series are zero.
Exercise 16.14
Consider a function f(x), deﬁned on 0 ≤ x ≤ L, which is odd around x =
L
2
.
Show that the even coeﬃcients (n even) of the Fourier cosine series are zero.
Exercise 16.15
The Fourier sine series of f(x) = cos
_
πx
L
_
for 0 < x < L is given by
cos
_
πx
L
_
=

n=1
b
n
sin
_
nπx
L
_
, n ∈ N
where
b
1
= 0, b
n
=
2n
(n
2
−1)π
[1 + (−1)
n
].
Using term-by-term integration, ﬁnd the Fourier cosine series of sin
_
nπx
L
_
.
16 FOURIER SINES SERIES AND FOURIER COSINES SERIES 141
Exercise 16.16
Consider the function
f(x) =
_
1 0 ≤ x < 1
2 1 ≤ x < 2
(a) Sketch the even extension of f.
(b) Find a
0
in the Fourier series for the even extension of f.
(c) Find a
n
(n = 1, 2, · · · ) in the Fourier series for the even extension of f.
(d) Find b
n
in the Fourier series for the even extension of f.
(e) Write the Fourier series for the even extension of f.
142SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
17 Separation of Variables for PDEs
Finding analytic solutions to PDEs is essentially impossible. Most of the
PDE techniques involve a mixture of analytic, qualitative and numeric ap-
proaches. Of course, there are some easy PDEs too. If you are lucky your
PDE has a solution with separable variables. In this section we discuss the
application of the method of separation of variables in the solution of PDEs.
In developing a solution to a partial diﬀerential equation by separation of
variables, one assumes that it is possible to separate the contributions of
the independent variables into separate functions that each involve only one
independent variable. Thus, the method consists of the following steps
1. Factorize the (unknown) dependent variable of the PDE into a product of
functions, each of the factors being a function of one independent variable.
That is,
u(x, y) = X(x)Y (y).
2. Substitute into the PDE, and divide the resulting equation by X(x)Y (y).
3. Then the problem turns into a set of separated ODEs (one for X(x) and
one for Y (y).)
4. The general solution of the ODEs is found, and boundary initial condi-
tions are imposed.
5. u(x, y) is formed by multiplying together X(x) and Y (y).
We illustrate these steps in the next two examples.
Example 17.1
Find all the solutions of the form u(x, t) = X(x)T(t) of the eqaution
u
xx
−u
x
= u
t
Solution.
It is very easy to ﬁnd the derivatives of a separable function:
u
x
= X

(x)T(t), u
t
= X(x)T

(t) and u
xx
= X

(x)T(t)
this is basically a consequence of the fact that diﬀerentiation with respect
to x sees t as a constant, and vice versa. Now the equation u
xx
− u
x
= u
t
becomes
X

(x)T(t) −X

(x)T(t) = X(x)T

(t).
17 SEPARATION OF VARIABLES FOR PDES 143
We can separate variables further. Division by X(x)T(t) gives
X

(x) −X

(x)
X(x)
=
T

(t)
T(t)
.
The expression on the LHS is a function of x whereas the one on the RHS is
a function on t only. They both have to be constant. That is,
X

(x) −X

(x)
X(x)
=
T

(t)
T(t)
= λ.
Thus, we have the following ODEs:
X

−X

−λX = 0 and T

= λT.
The second equation is easy to solve: T(t) = Ce
λt
. The ﬁrst equation is
solved via the characteristic equation ω
2
−ω −λ = 0, whose solutions are
ω =
1 ±

1 + 4λ
2
.
If λ > −
1
4
then
X(x) = Ae
1+

1+4λ
2
x
+ Be
1−

1+4λ
2
x
.
In this case,
u(x, t) = De
1+

1+4λ
2
x
e
λt
+ Ee
1−

1+4λ
2
x
e
λt
.
If λ = −
1
4
then
X(x) = Ae
x
2
+ Bxe
x
2
and in this case
u(x, t) = (D + Ex)e
x
2

t
4
.
If λ < −
1
4
then
X(x) = Ae
x
2
cos
_
_
−(1 + 4λ)
2
x
_
+ Be
x
2
sin
_
_
−(1 + 4λ)
2
x
_
.
In this case,
u(x, t) = De
x
2
+λt
cos
_
_
−(1 + 4λ)
2
x
_
+ Be
x
2
+λt
sin
_
_
−(1 + 4λ)
2
x
_
144SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
Example 17.2
Solve Laplace’s equation using the separation of variables method
∆u = u
xx
+ u
yy
= 0.
Solution.
We look for a solution of the form u(x, y) = X(x)Y (y). Substituting in the
Laplace’s equation, we obtain
X

(x)Y (y) + X(x)Y

(y) = 0.
Assuming X(x)Y (y) is nonzero, dividing for X(x)Y (y) and subtracting
Y

(y)
Y (y)
from both sides, we ﬁnd:
X

(x)
X(x)
= −
Y

(y)
Y (y)
.
The left hand side is a function of x while the right hand side is a function
of y. This says that they must equal to a constant. That is,
X

(x)
X(x)
= −
Y

(y)
Y (y)
= λ
where λ is a constant. This results in the following two ODEs
X

−λX = 0 and Y

+ λY = 0.
The solutions of these equations depend on the sign of λ.
• If λ > 0 then the solutions are given
X(x) =Ae

λx
+ Be

λx
Y (y) =C cos

λy + Dsin

λy
where A, B, C, and D are constants. In this case,
u(x, t) =k
1
e

λx
cos

λy + k
2
e

λx
sin

λy
+k
3
e

λx
cos

λy + k
4
e

λx
sin

λy.
• If λ = 0 then
X(x) =Ax + B
Y (y) =Cy + D
17 SEPARATION OF VARIABLES FOR PDES 145
where A, B, and C are arbitrary constants. In this case,
u(x, y) = k
1
xy + k
2
x + k
3
y + k
4
.
• If λ < 0 then
X(x) =Acos

−λx + Bsin

−λx
Y (y) =Ce

−λy
+ De

−λy
where A, B, C, and D are arbitrary constants. In this case,
u(x, y) =k
1
cos

−λxe

−λy
+ k
2
cos

−λxe

−λy
+k
3
sin

−λxe

−λy
+ k
4
sin

−λxe

−λy
Example 17.3
Solve using the separation of variables method.
yu
x
−xu
y
= 0.
Solution.
Substitute u(x, t) = X(x)Y (y) into the given equation we ﬁnd
yX

Y −xXY

= 0.
This can be separated into
X

xX
=
Y

yY
.
The left hand side is a function of x while the right hand side is a function
of y. This says that they must equal to a constant. That is,
X

xX
=
Y

yY
= λ.
where λ is a constant. This results in the following two ODEs
X

−λxX = 0 and Y

−λyY = 0.
Solving these equations using the method of separation of variable for ODEs
we ﬁnd X(x) = Ae
λx
2
2
and Y (y) = Be
λy
2
2
. Thus, the general solution is given
by
u(x, y) = Ce
λ(x
2
+y
2
)
2
146SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
Practice Problems
Exercise 17.1
Solve using the separation of variables method
∆u + λu = 0
Exercise 17.2
Solve using the separation of variables method
u
t
= ku
xx
.
Exercise 17.3
Derive the system of ordinary diﬀerential equations for X(x) and T(t) that
is satisﬁed by solutions to
u
rr
+
1
r
u
r
+
1
r
2
u
θθ
= 0.
Exercise 17.4
Derive the system of ordinary diﬀerential equations and boundary conditions
for X(x) and T(t) that is satisﬁed by solutions to
u
tt
= u
xx
−2u, 0 < x < 1, t > 0
u(0, t) = 0 = u
x
(1, t) t > 0
of the form u(x, t) = X(x)T(t). (Note: you do not need to solve for X and
T.)
Exercise 17.5
Derive the system of ordinary diﬀerential equations and boundary conditions
for X(x) and T(t) that is satisﬁed by solutions to
u
t
= ku
xx
, 0 < x < L, t > 0
u(x, 0) = f(x), u(0, t) = 0 = u
x
(L, t) t > 0
of the form u(x, t) = X(x)T(t). (Note: you do not need to solve for X and
T.)
Exercise 17.6
Find all product solutions of the PDE u
x
+ u
t
= 0.
17 SEPARATION OF VARIABLES FOR PDES 147
Exercise 17.7
Derive the system of ordinary diﬀerential equations for X(x) and Y (y) that
is satisﬁed by solutions to
3u
yy
−5u
xxxy
+ 7u
xxy
= 0.
of the form u(x, y) = X(x)Y (y).
Exercise 17.8
Find the general solution by the method of separation of variables.
u
xy
+ u = 0.
Exercise 17.9
Find the general solution by the method of separation of variables.
u
x
−yu
y
= 0.
148SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
Sample Exam Questions
Exercise 17.10
Find the general solution by the method of separation of variables.
u
tt
−u
xx
= 0.
Exercise 17.11
For the following PDEs ﬁnd the ODEs implied by the method of separation
of variables.
(a) u
t
= kr(ru
r
)
r
(b) u
t
= ku
xx
−αu
(c) u
t
= ku
xx
−au
x
(d) u
xx
+ u
yy
= 0
(e) u
t
= ku
xxxx
.
Exercise 17.12
Find all solutions to the following partial diﬀerential equation that can be
obtained via the separation of variables.
u
x
−u
y
= 0.
Exercise 17.13
Separate the PDE u
xx
−u
y
+u
yy
= u into two ODEs with a parameter. You
do not need to solve the ODEs.
18 SOLUTIONS OF THE HEAT EQUATIONBYTHE SEPARATIONOF VARIABLES METHOD149
18 Solutions of the Heat Equation by the Sep-
aration of Variables Method
In this section we apply the method of separation of variables in solving the
one spatial dimension of the heat equation.
The Heat Equation with Dirichlet Boundary Conditions
Consider the problem of ﬁnding all nontrivial solutions to the heat equation
u
t
= ku
xx
that satisﬁes the initial time condition u(x, 0) = f(x) and the
Dirichlet boundary conditions u(0, t) = T
0
and u(L, t) = T
L
.
From Exercise 14.4, it suﬃces to solve the problem with the Dirichlet bound-
ary conditions being replaced by the homogeneous boundary conditions u(0, t) =
u(L, t) = 0 (that is, the endpoints are assumed to be at zero temperature)
with u not the trivial solution. Let’s assume that the solution can be writ-
ten in the form u(x, t) = X(x)T(t). Substituting into the heat equation we
obtain
X

X
=
T

kT
.
Since the LHS only depends on x and the RHS only depends on t, then there
must be a constant λ such that
X

X
= λ and
T

kT
= λ.
This gives the two ordinary diﬀerential equations
X

−λX = 0 and T

−kλT = 0.
As far as the boundary conditions, we have
u(0, t) = 0 = X(0)T(t) =⇒X(0) = 0
and
u(L, t) = 0 = X(L)T(t) =⇒X(L) = 0.
Note that T is not the zero function for otherwise u ≡ 0 and this contradicts
our assumption that u is the non-trivial solution.
Next, we consider the three cases of the sign of λ.
150SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
Case 1: λ = 0
In this case, X

= 0. Solving this equation we ﬁnd X(x) = ax + b. Since
X(0) = 0 we ﬁnd b = 0. Since X(L) = 0 we ﬁnd a = 0. Hence, X ≡ 0 and
u(x, t) ≡ 0. That is, u is the trivial solution.
Case 2: λ > 0
In this case, X(x) = Ae

λx
+Be

λx
. Again, the conditions X(0) = X(L) =
0 imply A = B = 0 and hence the solution is the trivial solution.
Case 3: λ < 0
In this case, X(x) = Acos

−λx + Bsin

−λx. The condition X(0) = 0
implies A = 0. The condition X(L) = 0 implies Bsin

−λL = 0. We must
have B = 0 otherwise X(x) = 0 and this leads to the trivial solution. Since
B = 0, we obtain sin

−λL = 0 or

−λL = nπ where n ∈ N. Solving for λ
we ﬁnd λ = −
n
2
π
2
L
2
. Thus, we obtain inﬁnitely many solutions given by
X
n
(x) = A
n
sin

L
x, n ∈ N.
Now, solving the equation
T

−λkT = 0
by the method of separation of variables we obtain
T
n
(t) = B
n
e

n
2
π
2
L
2
kt
, n ∈ N.
Hence, the functions
u
n
(x, t) = C
n
sin
_

L
x
_
e

n
2
π
2
L
2
kt
, n ∈ N
satisfy u
t
= ku
xx
and the boundary conditions u(0, t) = u(L, t) = 0.
Now, in order for these solutions to satisfy the initial value condition u(x, 0) =
f(x), we invoke the superposition principle of linear PDE to write
u(x, t) =

n=1
C
n
sin
_

L
x
_
e

n
2
π
2
L
2
kt
. (18.1)
To determine the unknown constants C
n
we use the initial condition u(x, 0) =
f(x) in (18.1) to obtain
f(x) =

n=1
C
n
sin
_

L
x
_
.
18 SOLUTIONS OF THE HEAT EQUATIONBYTHE SEPARATIONOF VARIABLES METHOD151
Since the right-hand side is the Fourier sine series of f on the interval [0, L],
the coeﬃcients C
n
are given by
C
n
=
2
L
_
L
0
f(x) sin
_

L
x
_
dx. (18.2)
Thus, the solution to the heat equation is given by (18.1) with the C

n
s
calculated from (18.2).
Remark 18.1
According to Exercise 14.4, the solution to the heat equation with non-
homogeneous condition u(0, t) = T
0
and u(L, t) = T
L
is given by
u(x, t) =

n=1
C
n
sin
_

L
x
_
e

n
2
π
2
L
2
kt
+ T
0
+
T
L
−T
0
L
x.
The Heat Equation with Neumann Boundary Conditions
When both ends of the bar are insulated, that is, there is no heat ﬂow out
of them, we use the boundary conditions
u
x
(0, t) = u
x
(L, t) = 0.
In this case, the general form of the heat equation initial boundary value
problem is to ﬁnd u(x, t) satisfying
u
t
(x, t) =ku
xx
(x, t), 0 ≤ x ≤ L, t > 0
u(x, 0) =f(x), 0 ≤ x ≤ L
u
x
(0, t) =u
x
(L, t) = 0, t > 0.
Since 0 = u
x
(0, t) = X

(0)T(t) we obtain X

(0) = 0. Likewise, 0 = u
x
(L, t) =
X

(L)T(t) implis X

(L) = 0. Now, diﬀerentiating X(x) = Acos

−λx +
Bsin

−λx with respect to x we ﬁnd
X

(x) = −

−λAsin

−λx +

−λBcos

−λx.
The conditions X

(0) = X

(L) = 0 imply

−λB = 0 and

−λAsin

−λL =
0. Hence, B = 0 and λ = −
n
2
π
2
L
2
and
X
n
(x) = A
n
cos
_

L
x
_
, n = 0, 1, 2, · · ·
152SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
and
u
n
(x, t) = C
n
cos
_

L
x
_
e

n
2
π
2
L
2
kt
.
By the superposition principle, the required solution to the heat equation
with Neumann boundary conditions is given by
u(x, t) =

n=0
C
n
cos
_

L
x
_
e

n
2
π
2
L
2
kt
where
C
0
=
1
L
_
L
0
f(x)dx
C
n
=
2
L
_
L
0
f(x) cos
_

L
x
_
dx, n ∈ N.
18 SOLUTIONS OF THE HEAT EQUATIONBYTHE SEPARATIONOF VARIABLES METHOD153
Practice Problems
Exercise 18.1
Find the temperature in a bar of length 2 whose ends are kept at zero
and lateral surface insulated if the initial temperature is f(x) = sin
_
π
2
x
_
+
3 sin
_

2
x
_
.
Exercise 18.2
Find the temperature in a homogeneous bar of heat conducting material of
length L with its end points kept at zero and initial temperature distribution
given by f(x) =
dx
L
2
(L −x).
Exercise 18.3
Find the temperature in a thin metal rod of length L, with both ends insu-
lated (so that there is no passage of heat through the ends) and with initial
temperature in the rod f(x) = sin
_
π
L
x
_
.
Exercise 18.4
Solve the following heat equation with Dirichlet boundary conditions
u
t
= ku
xx
u(0, t) = u(L, t) = 0
u(x, 0) =
_
1 0 < x <
L
2
2
L
2
≤ x < L
Exercise 18.5
Solve
u
t
= ku
xx
u(0, t) = u(L, t) = 0
u(x, 0) = 6 sin
_

L
x
_
.
Exercise 18.6
Solve
u
t
= ku
xx
subject to
u
x
(0, t) = u
x
(L, t) = 0
u(x, 0) =
_
0 0 < x <
L
2
1
L
2
≤ x < L
154SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
Exercise 18.7
Solve
u
t
= ku
xx
subject to
u
x
(0, t) = u
x
(L, t) = 0
u(x, 0) = 6 + 4 cos
_

L
x
_
.
Exercise 18.8
Solve
u
t
= ku
xx
subject to
u
x
(0, t) = u
x
(L, t) = 0
u(x, 0) = −3 cos
_

L
x
_
.
Exercise 18.9
Find the general solution of the following equation and brieﬂy describe its
limit as t →∞.
u
t
= u
xx
, 0 < x < L, t ∈ R
u
x
(0, t) = 0 = u(L, t), t ∈ R.
Exercise 18.10
Find the general solution u(x, t) of
u
t
= u
xx
−u, 0 < x < L, t > 0
u
x
(0, t) = 0 = u
x
(L, t), t > 0.
Brieﬂy describe its behavior as t →∞.
18 SOLUTIONS OF THE HEAT EQUATIONBYTHE SEPARATIONOF VARIABLES METHOD155
Sample Exam Questions
Exercise 18.11 (Energy method)
Let u
1
and u
2
be two solutions to the Robin boundary value problem
u
t
= u
xx
−u, 0 < x < 1, t > 0
u
x
(0, t) = u
x
(1, t) = 0, t > 0
u(x, 0) = g(x), 0 < x < 1
Deﬁne w(x, t) = u
1
(x, t) −u
2
(x, t).
(a) Show that w satisﬁes the initial value problem
w
t
= w
xx
−w, 0 < x < 1, t > 0
w(x, 0) = 0, 0 < x < 1
(b) Deﬁne E(t) =
_
1
0
w
2
(x, t)dx ≥ 0 for all t ≥ 0. Show that E

(t) ≤ 0.
Hence, 0 ≤ E(t) ≤ E(0) for all t > 0.
(c) Show that E(t) = 0, w(x, t) = 0. Hence, conclude that u
1
= u
2
.
Exercise 18.12
Consider the heat induction in a bar where the left end temperature is main-
tained at 0, and the right end is perfectly insulated. We assume k = 1 and
L = 1.
(a) Derive the boundary conditions of the temperature at the endpoints.
(b) Following the separation of variables approach, derive the ODEs for X
and T.
(c) Consider the equation in X(x). What are the values of X(0) and X(1)?
Show that solutions of the form X(x) = sin

−λx satisﬁy the ODE and one
of the boundary conditions. Can you choose a value of λ so that the other
boundary condition is also satisﬁed?
Exercise 18.13
Using the method of separation of variables ﬁnd the solution of the heat
equation
u
t
= ku
xx
satisfying the following boundary and initial conditions:
(a) u(0, t) = u(L, t) = 0, u(x, 0) = 6 sin
_
9πx
L
_
(b) u(0, t) = u(L, t) = 0, u(x, 0) = 3 sin
_
πx
L
_
−sin
_
3πx
L
_
156SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
Exercise 18.14
Using the method of separation of variables ﬁnd the solution of the heat
equation
u
t
= ku
xx
satisfying the following boundary and initial conditions:
(a) u
x
(0, t) = u
x
(L, t) = 0, u(x, 0) = cos
_
πx
L
_
+ 4 cos
_
5πx
L
_
.
(b) u
x
(0, t) = u
x
(L, t) = 0, u(x, 0) = 5.
Exercise 18.15
Find the solution of the following heat conduction partial diﬀerential equation
u
t
= 8u
xx
, 0 < x < 4π, t > 0
u(0, t) = u(4π, t) = 0, t > 0
u(x, 0) = 6 sin x, 0 < x < 4π.
19 ELLIPTIC TYPE: LAPLACE’S EQUATIONS INRECTANGULAR DOMAINS157
19 Elliptic Type: Laplace’s Equations in Rect-
angular Domains
Boundary value problems are of great importance in physical applications.
Mathematically, a boundary-value problem consists of ﬁnding a function
which satisﬁes a given partial diﬀerential equation and particular bound-
ary conditions. Physically speaking, the problem is independent of time,
involving only space coordinates.
Just as initial-value problems are associated with hyperbolic PDE, bound-
ary value problems are associated with PDE of elliptic type. In contrast to
initial-value problems, boundary-value problems are considerably more diﬃ-
cult to solve.
The main model example of an elliptic type PDE is the Laplace equation
∆u = u
xx
+ u
yy
= 0. (19.1)
where the symbol ∆ is referred to as the Laplacian. Solutions of this equa-
tion are called harmonic functions.
Example 19.1
Show that, for all (x, y) = (0, 0), u(x, y) = ln (x
2
+ y
2
) is a harmonic function.
Solution.
We have
u
x
=
2x
x
2
+ y
2
u
xx
=
2y
2
−2x
2
(x
2
+ y
2
)
2
u
y
=
2y
x
2
+ y
2
u
yy
=
2x
2
−2y
2
(x
2
+ y
2
)
2
Plugging these expressions into the equation we ﬁnd u
xx
+ u
yy
= 0. Hence,
u(x, y) is harmonic
The Laplace equation is arguably the most important diﬀerential equation in
158SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
all of applied mathematics. It arises in an astonishing variety of mathemati-
cal and physical systems, ranging through ﬂuid mechanics, electromagnetism,
potential theory, solid mechanics, heat conduction, geometry, probability,
number theory, and on and on.
There are two main modiﬁcations of the Laplace equation: the Poisson
equation (a non-homogeneous Laplace equation):
∆u = f(x, y)
and the eigenvalue problem (the Helmholtz equation):
∆u = λu, λ ∈ R.
Solving Laplace’s Equation (19.1)
Note ﬁrst that both independent variables are spatial variables and each
variable occurs in a 2nd order derivative and so we will need two boundary
conditions for each variable a total of four boundary conditions.
Consider (19.1) in the rectangle
Ω = {(x, y) : 0 ≤ x ≤ a, 0 ≤ y ≤ b}
with the Dirichlet boundary conditions
u(0, y) = f
1
(y), u(a, y) = f
2
(y), u(x, 0) = g
1
(x), u(x, b) = g
2
(x)
where 0 ≤ x ≤ a and 0 ≤ y ≤ b.
The separation of variables method is most successful when the boundary
conditions are homogeneous. Thus, solving the Laplace’s equation in Ω re-
quires solving four initial boundary conditions problems, where in each prob-
lem three of the four conditions are homogeneous. The four problems to be
solved are
(I)
_
_
_
u
xx
+u
yy
= 0
u(0, y) = f
1
(y),
u(a, y) = u(x, 0) = u(x, b) = 0
(II)
_
_
_
u
xx
+u
yy
= 0
u(a, y) = f
2
(y),
u(0, y) = u(x, 0) = u(x, b) = 0
(III)
_
_
_
u
xx
+u
yy
= 0
u(x, 0) = g
1
(x),
u(0, y) = u(a, y) = u(x, b) = 0
(IV )
_
_
_
u
xx
+u
yy
= 0
u(x, b) = g
2
(x),
u(0, y) = u(a, y) = u(x, 0) = 0
19 ELLIPTIC TYPE: LAPLACE’S EQUATIONS INRECTANGULAR DOMAINS159
If we let u
i
(x, y), i = 1, 2, 3, 4, denote the solution of each of the above
problems, then the solution to our original system will be
u(x, y) = u
1
(x, y) + u
2
(x, y) + u
3
(x, y) + u
4
(x, y).
In each of the above problems, we will apply separation of variables to (19.1)
and ﬁnd a product solution that will satisfy the diﬀerential equation and the
three homogeneous boundary conditions. Using the Principle of Superposi-
tion we will ﬁnd a solution to the problem and then apply the ﬁnal boundary
condition to determine the value of the constant(s) that are left in the prob-
lem. The process is nearly identical in many ways to what we did when we
were solving the heat equation.
We will illustrate how to ﬁnd u(x, y) = u
4
(x, y). So let’s assume that the so-
lution can be written in the form u(x, y) = X(x)Y (y). Substituting in (19.1),
we obtain
X

(x)Y (y) + X(x)Y

(y) = 0.
Assuming X(x)Y (y) is nonzero, dividing for X(x)Y (y) and subtracting
Y

(y)
Y (y)
from both sides, we ﬁnd:
X

(x)
X(x)
= −
Y

(y)
Y (y)
.
The left hand side is a function of x while the right hand side is a function
of y. This says that they must equal to a constant. That is,
X

(x)
X(x)
= −
Y

(y)
Y (y)
= λ
where λ is a constant. This results in the following two ODEs
X

−λX = 0 and Y

+ λY = 0.
As far as the boundary conditions, we have for all 0 ≤ x ≤ a and 0 ≤ y ≤ b
u(0, y) = 0 = X(0)Y (y) =⇒X(0) = 0
u(a, y) = 0 = X(a)Y (y) =⇒X(a) = 0
u(x, 0) = 0 = X(x)Y (0) =⇒Y (0) = 0
u(x, b) = g
2
(x) = X(x)Y (b).
160SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
Note that X and Y are not the zero functions for otherwise u ≡ 0 and this
contradicts our assumption that u is the non-trivial solution.
Consider the ﬁrst equation: since X

−λX = 0 the solution depends on the
sign of λ. If λ = 0 then X(x) = Ax+B. Now, the conditions X(0) = X(a) = 0
imply A = B = 0 and so u ≡ 0. So assume that λ = 0. If λ > 0 then
X(x) = Ae

λx
+ Be

λx
. Now, the conditions X(0) = X(a) = 0, λ = 0
imply A = B = 0 and hence the solution is the trivial solution. Hence, in
order to have a nontrivial solution we must have λ < 0. In this case,
X(x) = Acos

−λx + Bsin

−λx.
The condition X(0) = 0 implies A = 0. The condition X(a) = 0 implies
Bsin

−λa = 0. We must have B = 0 otherwise X(x) = 0 and this leads to
the trivial solution. Since B = 0, we obtain sin

−λa = 0 or

−λa = nπ
where n ∈ N. Solving for λ we ﬁnd λ
n
= −
n
2
π
2
a
2
. Thus, we obtain inﬁnitely
many solutions given by
X
n
(x) = sin

a
x, n ∈ N.
Now, solving the equation
Y

+ λY = 0
we obtain
Y
n
(y) = a
n
e

−λny
+ b
n
e

−λny
= A
n
cosh
_
−λ
n
y + B
n
sinh
_
−λ
n
y, n ∈ N.
Using the boundary condition Y (0) = 0 we obtain A
n
= 0 for all n ∈ N.
Hence, the functions
u
n
(x, y) = B
n
sin
_

a
x
_
sinh
_

a
y
_
, n ∈ N
satisfy (19.1) and the boundary conditions u(0, y) = u(a, y) = u(x, 0) = 0.
Now, in order for these solutions to satisfy the boundary value condition
u(x, b) = g
2
(x), we invoke the superposition principle of linear PDE to write
u(x, y) =

n=1
B
n
sin
_

a
x
_
sinh
_

a
y
_
. (19.2)
To determine the unknown constants B
n
we use the boundary condition
u(x, b) = g
2
(x) in (19.2) to obtain
g
2
(x) =

n=1
_
B
n
sinh
_

a
b
__
sin
_

a
x
_
.
19 ELLIPTIC TYPE: LAPLACE’S EQUATIONS INRECTANGULAR DOMAINS161
Since the right-hand side is the Fourier sine series of f on the interval [0, a],
the coeﬃcients B
n
are given by
B
n
=
_
2
a
_
a
0
g
2
(x) sin
_

a
x
_
dx
_
[sinh
_

a
b
_
]
−1
. (19.3)
Thus, the solution to the Laplace’s equation is given by (19.1) with the B

n
s
calculated from (19.3).
Example 19.2
Solve
_
_
_
u
xx
+ u
yy
= 0
u(0, y) = f
1
(y),
u(a, y) = u(x, 0) = u(x, b) = 0
Solution.
Assume that the solution can be written in the form u(x, y) = X(x)Y (y).
Substituting in (19.1), we obtain
X

(x)Y (y) + X(x)Y

(y) = 0.
Assuming X(x)Y (y) is nonzero, dividing for X(x)Y (y) and subtracting
Y

(y)
Y (y)
from both sides, we ﬁnd:
X

(x)
X(x)
= −
Y

(y)
Y (y)
.
The left hand side is a function of x while the right hand side is a function
of y. This says that they must equal to a constant. That is,
X

(x)
X(x)
= −
Y

(y)
Y (y)
= λ
where λ is a constant. This results in the following two ODEs
X

−λX = 0 and Y

+ λY = 0.
As far as the boundary conditions, we have for all 0 ≤ x ≤ a and 0 ≤ y ≤ b
u(0, y) = f
1
(y) = X(0)Y (y)
u(a, y) = 0 = X(a)Y (y) =⇒X(a) = 0
162SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
u(x, 0) = 0 = X(x)Y (0) =⇒Y (0) = 0
u(x, b) = 0 = X(x)Y (b) =⇒Y (b) = 0
Note that X and Y are not the zero functions for otherwise u ≡ 0 and this
contradicts our assumption that u is the non-trivial solution.
Consider the second equation: since Y

+λY = 0 the solution depends on the
sign of λ. If λ = 0 then Y (y) = Ay+B. Now, the conditions Y (0) = Y (b) = 0
imply A = B = 0 and so u ≡ 0. So assume that λ = 0. If λ < 0 then
Y (y) = Ae

−λy
+ Be

−λy
. Now, the condition Y (0) = Y (b) = 0 imply
A = B = 0 and hence the solution is the trivial solution. Hence, in order to
have a nontrivial solution we must have λ > 0. In this case,
Y (y) = Acos

λy + Bsin

λy.
The condition Y (0) = 0 implies A = 0. The condition Y (b) = 0 implies
Bsin

λb = 0. We must have B = 0 otherwise Y (y) = 0 and this leads to
the trivial solution. Since B = 0, we obtain sin

λb = 0 or

λb = nπ where
n ∈ N. Solving for λ we ﬁnd λ
n
=
n
2
π
2
b
2
. Thus, we obtain inﬁnitely many
solutions given by
Y
n
(y) = sin
_

b
y
_
, n ∈ N.
Now, solving the equation
X

−λX = 0, λ > 0
we obtain
X
n
(x) = a
n
e

λnx
+ b
n
e

λnx
= A
n
cosh
_

b
x
_
+ B
n
sinh
_

b
x
_
, n ∈ N.
However, this is not really suited for dealing with the boundary condition
X(a) = 0. So, let’s also notice that the following is also a solution.
X
n
(x) = A
n
cosh
_

b
(x −a)
_
+ B
n
sinh
_

b
(x −a)
_
, n ∈ N.
Now, using the boundary condition X(a) = 0 we obtain A
n
= 0 for all n ∈ N.
Hence, the functions
u
n
(x, y) = B
n
sin
_

b
y
_
sinh
_

b
(x −a)
_
, n ∈ N
19 ELLIPTIC TYPE: LAPLACE’S EQUATIONS INRECTANGULAR DOMAINS163
satisfy (19.1) and the boundary conditions u(a, y) = u(x, 0) = u(x, b) = 0.
Now, in order for these solutions to satisfy the boundary value condition
u(0, y) = f
1
(y), we invoke the superposition principle of linear PDE to write
u(x, y) =

n=1
B
n
sin
_

b
y
_
sinh
_

b
(x −a)
_
. (19.4)
To determine the unknown constants B
n
we use the boundary condition
u(0, y) = f
1
(y) in (19.4) to obtain
f
1
(y) =

n=1
_
B
n
sinh
_

b
a
__
sin
_

b
y
_
.
Since the right-hand side is the Fourier sine series of f
1
on the interval [0, b],
the coeﬃcients B
n
are given by
B
n
=
_
2
b
_
b
0
f
1
(y) sin
_

b
y
_
dy
_
_
sinh −
_

b
a
__
−1
. (19.5)
Thus, the solution to the Laplace’s equation is given by (19.4) with the B

n
s
calculated from (19.5)
Example 19.3
Solve
u
xx
+ u
yy
= 0, 0 < x < L, 0 < y < H
u(0, y) = u(L, y) = 0, 0 < g < H
u(x, 0) = u
y
(x, 0), u(x, H) = f(x), 0 < x < L.
Solution.
Using separation of variables we ﬁnd
X

X
= −
Y

Y
= λ.
We ﬁrst solve
_
X

−λX = 0 0 < x < L
X(0) = X(L) = 0
We ﬁnd λ
n
= −
n
2
π
2
L
2
and
X
n
(x) = sin

L
x, n ∈ N.
164SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
Next we need to solve
_
Y

+ λY = 0 0 < y < H
Y (0) −Y

(0) = 0
The solution of the ODE is
Y
n
(y) = A
n
cosh
_

L
y
_
+ B
n
sinh
_

L
y
_
y, n ∈ N.
The boundary condition Y (0) −Y

(0) = 0 implies
A
n
−B
n

L
= 0.
Hence,
Y
n
= B
n

L
cosh
_

L
y
_
+ B
n
sinh
_

L
y
_
y, n ∈ N.
Using the superposition principle and the results above we have
u(x, y) =

n=1
B
n
sin

L
x
_

L
cosh
_

L
y
_
+ sinh
_

L
y
__
.
Substituting in the condition u(x, H) = f(x) we ﬁnd
f(x) =

n=1
B
n
sin

L
x
_

L
cosh
_

L
H
_
+ sinh
_

L
H
__
.
Recall the Fourier sine series of f on [0, L] given by
f(x) =

n=1
A
n
sin

L
x
where
A
n
=
2
L
_
L
0
f(x) sin
_

L
x
_
dx.
Thus, the general solution is given by
u(x, y) =

n=1
B
n
sin

L
x
_

L
cosh
_

L
y
_
+ sinh
_

L
y
__
.
with the B
n
satisfying
B
n
_

L
cosh
_

L
H
_
+ sinh
_

L
H
__
=
2
L
_
L
0
f(x) sin
_

L
x
_
dx
19 ELLIPTIC TYPE: LAPLACE’S EQUATIONS INRECTANGULAR DOMAINS165
Practice Problems
Exercise 19.1
Solve
_
_
_
u
xx
+ u
yy
= 0
u(a, y) = f
2
(y),
u(0, y) = u(x, 0) = u(x, b) = 0
Exercise 19.2
Solve
_
_
_
u
xx
+ u
yy
= 0
u(x, 0) = g
1
(x),
u(0, y) = u(a, y) = u(x, b) = 0
Exercise 19.3
Solve
_
_
_
u
xx
+ u
yy
= 0
u(x, 0) = u(0, y) = 0,
u(1, y) = 2y, u(x, 1) = 3 sin πx + 2x
where 0 ≤ x ≤ 1 and 0 ≤ y ≤ 1. Hint: Deﬁne U(x, y) = u(x, y) −2xy.
Exercise 19.4
Show that u(x, y) = x
2
−y
2
and u(x, y) = 2xy are harmonic functions.
Exercise 19.5
Solve
u
xx
+ u
yy
= 0, 0 ≤ x ≤ L, −
H
2
≤ y ≤
H
2
subject to
u(0, y) = u(L, y) = 0, −
H
2
< y <
H
2
u(x, −
H
2
) = f
1
(x), u(x,
H
2
) = f
2
(x), 0 ≤ x ≤ L.
Exercise 19.6
Consider a complex valued function f(z) = u(x, y)+iv(x, y) where i =

−1.
We say that f is holomorphic or analytic if and only if f can be expressed
as a power series in z, i.e.
u(x, y) + iv(x, y) =

n=0
a
n
z
n
.
(a) By diﬀerentiating with respect to x and y show that
166SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
u
x
= v
y
and u
y
= −v
x
These are known as the Cauchy-Riemann equations.
(b) Show that ∆u = 0 and ∆v = 0.
Exercise 19.7
Show that Laplace’s equation in polar coordinates is given by
u
rr
+
1
r
u
r
+
1
r
2
u
θθ
= 0.
Exercise 19.8
Solve
u
xx
+ u
yy
= 0, 0 ≤ x ≤ 2, 0 ≤ y ≤ 3
subject to
u(x, 0) = 0, u(x, 3) =
x
2
u(0, y) = sin
_

3
y
_
, u(2, y) = 7.
Exercise 19.9
Solve
u
xx
+ u
yy
= 0, 0 ≤ x ≤ L, 0 ≤ y ≤ H
subject to
u
y
(x, 0) = 0, u(x, H) = 0
u(0, y) = u(L, y) = 4 cos
_
πy
2H
_
.
Exercise 19.10
Solve
u
xx
+ u
yy
= 0, x > 0, 0 ≤ y ≤ H
subject to
u(0, y) = f(y), |u(x, 0)| < ∞
u
y
(x, 0) = u
y
(x, H) = 0.
19 ELLIPTIC TYPE: LAPLACE’S EQUATIONS INRECTANGULAR DOMAINS167
Sample Exam Questions
Exercise 19.11
Consider Laplace’s equation inside a rectangle
u
xx
+ u
yy
= 0, 0 ≤ x ≤ L, 0 ≤ y ≤ H
subject to the boundary conditions
u(0, y) = 0, u(L, y) = 0, u(x, 0)−u
y
(x, 0) = 0, u(x, H) = 20 sin
_
πx
L
_
−5 sin
_
3πx
L
_
.
Find the solution u(x, y).
Exercise 19.12
Solve Laplace’e equation u
xx
+ u
yy
= 0 in the rectangle 0 < x, y < 1 subject
to the conditions
u(0, y) = u(1, y) = 0, 0 < y < 1
u(x, 0) = sin (2πx), u
x
(x, 0) = −2π sin (2πx), 0 < x < 1.
Exercise 19.13
Find the solution to Laplace’s equation on the rectangle 0 < x < 1, 0 < y < 1
with boundary conditions
u(x, 0) = 0, u(x, 1) = 1
u
x
(0, y) = u
x
(1, y) = 0.
Exercise 19.14
Solve Laplace’s equation on the rectangle 0 < x < a, 0 < y < b with the
boundary conditions
u
x
(0, y) = −a, u
x
(a, y) = 0
u
y
(x, 0) = b, u
y
(x, b) = 0.
Exercise 19.15
Solve Laplace’s equation on the rectangle 0 < x < π, 0 < y < 2 with the
boundary conditions
u(0, y) = u(π, y) = 0
u
y
(x, 0) = 0, u
y
(x, 2) = 2 sin 3x −5 sin 10x.
168SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
20 Laplace’s Equations in Circular Regions
In the previous section we solved the Dirichlet problem for Laplace’s equation
on a rectangular region. However, if the domain of the solution is a disc,
an annulus, or a circular wedge, it is useful to study the two-dimensional
Laplace’s equation in polar coordinates.
It is well known in calculus that the cartesian coordinates (x, y) and the polar
coordinates (r, θ) of a point are related by the formulas
x = r cos θ and y = r sin θ
where r = (x
2
+ y
2
)
1
2
and tan θ =
y
x
. Using the chain rule we obtain
u
x
=u
r
r
x
+ u
θ
θ
x
= cos θu
r

sin θ
r
u
θ
u
xx
=u
xr
r
x
+ u

θ
x
=
_
cos θu
rr
+
sin θ
r
2
u
θ

sin θ
r
u

_
cos θ
+
_
−sin θu
r
+ cos θu

cos θ
r
u
θ

sin θ
r
u
θθ
__

sin θ
r
_
u
y
=u
r
r
y
+ u
θ
θ
y
= sin θu
r
+
cos θ
r
u
θ
u
yy
=u
yr
r
y
+ u

θ
y
=
_
sin θu
rr

cos θ
r
2
u
θ
+
cos θ
r
u

_
sin θ
+
_
cos θu
r
+ sin θu

sin θ
r
u
θ
+
cos θ
r
u
θθ
__
cos θ
r
_
Substituting these equations into ∆u = 0 we obtain
u
rr
+
1
r
u
r
+
1
r
2
u
θθ
= 0. (20.1)
Example 20.1
Find the solution to
∆u = 0, x
2
+ y
2
< a
2
subject to
u(x, y) = f(θ), x
2
+ y
2
= a
2
.
20 LAPLACE’S EQUATIONS IN CIRCULAR REGIONS 169
Solution.
Two important facts that play an important role in the solution of the prob-
lem must be taken into consideration, namely:
(1) u(r, θ) must be bounded at r = 0.
(2) u(r, θ + 2π) = u(r, θ), that is, u is periodic of period 2π.
Now, we will apply the method of separation of variables to (21.1). Suppose
that a solution u(r, θ) of (21.1) can be written in the form u(r, θ) = R(r)Θ(θ).
Substituting in (21.1) we obtain
R

(r)Θ(θ) +
1
r
R

(r)Θ(θ) +
1
r
2
R(r)Θ

(θ) = 0
Dividing by RΘ (under the assumption that RΘ = 0) we obtian
Θ

(θ)
Θ(θ)
= −r
2
R

(r)
R(r)
−r
R

(r)
R(r)
.
The left-hand side is independent of r whereas the right-hand side is inde-
pendent of θ so that there is a constant λ such that

Θ

(θ)
Θ(θ)
= r
2
R

(r)
R(r)
+ r
R

(r)
R(r)
= λ.
This results in the following ODEs
Θ

(θ) + λΘ(θ) = 0 (20.2)
and
r
2
R

(r) + rR

(r) −λR(r) = 0 (20.3)
The second equation is known as Euler’s equation. Both of these equations
are easily solvable. We only have to add the appropriate boundary conditions.
By (2) above we must have
Θ(θ + 2π) = Θ(θ), ∀θ ∈ R.
In particular, we have Θ(0) = Θ(2π) and Θ

(0) = Θ

(2π). The periodicity of
Θ implies that λ = n
2
and Θ must be of the form
Θ
n
(θ) = A

n
cos nθ + B

n
sin nθ, n = 0, 1, 2 · · ·
170SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
The equation in R is of Euler type and its solution must be of the form
R(r) = r
α
. Since λ = n
2
, the corresponding characteristic equation is
α(α −1)r
α
+ αr
α
−n
2
r
α
= 0.
Solving this equation we ﬁnd α = ±n. Hence, we let
R
n
(r) = C
n
r
n
+ D
n
r
−n
, n ∈ N.
For n = 0, R = 1 is a solution. To ﬁnd a second solution, we solve the
equation
r
2
R

+ rR

= 0.
This can be done by dividing through by r and using the substitution S = R

to obtain rS

+ S = 0. Solving this by noting that the left-hand side is just
(rS)

we ﬁnd S =
c
r
. Hence, R

=
c
r
and this implies R(r) = C ln r. Thus,
R = 1 and R = ln r form a couple of linearly independent solutions of (20.3)
and so a general solution is given by
R
0
(r) = C
0
+ D
0
ln r.
By assumption (1), u(r, θ) must be bounded at r = 0, and so does R
n
. Since
r
−n
and ln r are unbounded at r = 0, we must set D
0
= D
n
= 0. In this case,
the solutions to Euler’s equation are given by
R
n
(r) = C
n
r
n
, n = 0, 1, 2, · · · .
Using the superposition principle, and combining the results obtained above,
we ﬁnd
u(r, θ) = C
0
+

n=1
r
n
(A
n
cos nθ + B
n
sin nθ).
Now, using the boundary condition u(a, θ) = f(θ) we can write
f(θ) = C
0
+

n=1
(a
n
A
n
cos nθ + a
n
B
n
sin nθ)
which is usually written in a more convenient equivalent form by
f(θ) =
a
0
2
+

n=1
(a
n
cos nθ + b
n
sin nθ).
20 LAPLACE’S EQUATIONS IN CIRCULAR REGIONS 171
It is obvious that a
n
and b
n
are the Fourier coeﬃcients, and therefore can be
determined by the formulas
a
n
=
1
π
_

0
f(θ) cos nθdθ, n = 0, 1, · · ·
and
b
n
=
1
π
_

0
f(θ) sin nθdθ, n = 1, 2, · · · .
Finally, the general solution to our problem is given by
u(r, θ) = C
0
+

n=1
r
n
(A
n
cos nθ + B
n
sin nθ)
where
C
0
=
a
0
2
=
1

_

0
f(θ)dθ
A
n
=
a
n
a
n
=
1
a
n
π
_

0
f(θ) cos nθdθ, n = 1, 2, · · ·
B
n
=
b
n
a
n
=
1
a
n
π
_

0
f(θ) sin nθdθ, n = 1, 2, · · ·
Example 20.2
Solve
∆u = 0, 0 ≤ θ < 2π, 1 ≤ r ≤ 2
subject to
u(1, θ) = u(2, θ) = sin θ, 0 ≤ θ < 2π.
Solution.
Use separation of variables. First, solving for Θ(θ)), we see that in order
to ensure that the solution is 2π−periodic in θ, the eigenvalues are λ = n
2
.
When solving the equation for R(r), we do NOT need to throw out solutions
which are not bounded as r → 0. This is because we are working in the
annulus where r is bounded away from 0 and ∞. Therefore, we obtain the
general solution
u(r, θ) = (C
0
+C
1
ln r) +

n=1
[(C
n
r
n
+D
n
r
−n
) cos nθ +(A
n
r
n
+B
n
r
−n
) sin nθ].
172SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
But
C
0
+

n=1

n=1
[(C
n
+ D
n
) cos nθ + (A
n
+ B
n
) sin nθ] = sin θ
and
C
0
+

n=1

n=1
[(C
n
2
n
+ D
n
2
−n
) cos nθ + (A
n
2
n
+ B
n
2
−n
) sin nθ] = sin θ
Hence, comparing coeﬃcients we must have
C
0
=0
C
n
+ D
n
=0
A
n
+ B
n
=0 n = 1
A
1
+ B
1
=1
C
n
2
n
+ D
n
2
−n
=0
A
n
2
n
+ B
n
2
−n
=0 n = 1
2A
1
+ 2
−1
B
1
=0
Solving these equations we ﬁnd C
0
= C
n
= D
n
= 0, A
1
=
1
3
, B
1
=
2
3
, and
A
n
= B
n
= 0 for n = 1. Hence, the solution to the problem is
u(r, θ) =
1
3
_
r +
2
r
_
sin θ
Example 20.3
Solve Laplace’s equation inside a 60

wedge of radius a subject to the bound-
ary conditions
u
θ
(r, θ) = 0, u
θ
(r,
π
3
) = 0, u(a, θ) =
1
3
cos 9θ −
1
9
cos 3θ.
You may assume that the solution remains bounded as r →0.
Solution.
Separating the variables we obtain the eigenvalue problem
Θ

(θ) + λΘ(θ) = 0
Θ

(0) = Θ

_
π
3
_
= 0.
20 LAPLACE’S EQUATIONS IN CIRCULAR REGIONS 173
As above, because of periodicity we expect the solution to be of the form
Θ(θ) = Acos

λθ + Bsin

λθ.
The condition Θ

(0) = 0 implies A = 0. The condition Θ

_
π
3
_
= 0 implies
λ
n
= (3n)
2
, n = 0, 1, 2, · · · . Thus, the angular solution is
Θ
n
(θ) = cos 3nθ, n = 0, 1, 2, · · ·
The corresponding solutions of the radial problem are
R
n
(r) = A
n
r
3n
+ B
n
r
3
n
, n = 0, 1, · · · .
To obtain a solution that remains bounded as r →0 we take B
n
= 0. Hence,
u
n
(r, θ) =

n=0
C
n
r
3n
cos 3nθ, n = 0, 1, 2, · · ·
Using the boundary condition
u(a, θ) =
1
3
cos 9θ −
1
9
cos 3θ
we obtain C
1
a
3
= −
1
9
and C
3
a
9
=
1
3
and 0 otherwise. Thus,
u(a, θ) =
1
3
_
r
a
_
9
cos 9θ −
1
9
_
r
a
_
3
cos 3θ
174SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
Practice Problems
Exercise 20.1
Solve the Laplace’s equation in the unit disk with u(1, θ) = 3 sin 5θ.
Exercise 20.2
Solve the Laplace’s equation in the upper half of the unit disk with u(1, θ) =
π −θ.
Exercise 20.3
Solve the Laplace’s equation in the unit disk with u
r
(1, θ) = 2 cos 2θ.
Exercise 20.4
Consider
u(r, θ) = C
0
+

n=1
r
n
(A
n
cos nθ + B
n
sin nθ)
with
C
0
=
a
0
2
=
1

_

0
f(φ)dφ
A
n
=
a
n
a
n
=
1
a
n
π
_

0
f(φ) cos nφdφ, n = 1, 2, · · ·
B
n
=
b
n
a
n
=
1
a
n
π
_

0
f(φ) sin nφdφ, n = 1, 2, · · ·
Using the trigonometric identity
cos a cos b + sin a sin b = cos (a −b)
show that
u(r, θ) =
1

_

0
f(φ)
_
1 + 2

n=1
_
r
a
_
n
cos n(θ −φ)
_
dφ.
Exercise 20.5
(a) Using Euler’s formula from complex analysis e
it
= cos t +i sin t show that
cos t =
1
2
(e
it
+ e
−it
),
20 LAPLACE’S EQUATIONS IN CIRCULAR REGIONS 175
where i =

−1.
(b) Show that
1 + 2

n=1
_
r
a
_
n
cos n(θ −φ) = 1 +

n=1
_
r
a
_
n
e
in(θ−φ)
+

n=1
_
r
a
_
n
e
−in(θ−φ)
.
(c) Let q
1
=
r
a
e
i(θ−φ)
and q
2
=
r
a
e
−i(θ−φ)
. It is deﬁned in complex analysis that
the absolute value of a complex number z = x+iy is given by |z| = (x
2
+y
2
)
1
2
.
Using these concepts, show that |q
1
| < 1 and |q
2
| < 1.
Exercise 20.6
(a)Show that

n=1
_
r
a
_
n
e
in(θ−φ)
=
re
i(θ−φ)
a −re
i(θ−φ)
and

n=1
_
r
a
_
n
e
−in(θ−φ)
=
re
−i(θ−φ)
a −re
−i(θ−φ)
Hint: Each sum is a geoemtric series with a ratio less than 1 in absolute
value so that these series converges.
(b) Show that
1 + 2

n=1
_
r
a
_
n
cos n(θ −φ) =
a
2
−r
2
a
2
−2ar cos (θ −φ) + r
2
.
Exercise 20.7
Show that
u(r, θ) =
a
2
−r
2

_

0
f(φ)
a
2
−2ar cos (θ −φ) + r
2
dφ.
This is known as the Poisson formula in polar coordinates.
Exercise 20.8
Solve
u
xx
+ u
yy
= 0, x
2
+ y
2
< 1
subject to
u(1, θ) = θ, −π ≤ θ ≤ π.
176SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
Exercise 20.9
The vibrations of a symmetric circular membrane where the displacement
u(r, t) depends on r and t only can be describe by the one-dimensional wave
equation in polar coordinates
u
tt
= c
2
(u
rr
+
1
r
u
r
), 0 < r < a, t > 0
with initial condition
u(a, t) = 0, t > 0
and boundary conditions
u(r, 0) = f(r), u
t
(r, 0) = g(r), 0 < r < a.
(a) Show that the assumption u(r, t) = R(r)T(t) leads to the equation
1
c
2
T

T
=
1
R
R

+
1
r
R

R
= λ.
(b) Show that λ < 0.
Exercise 20.10
Cartesian coordinates and cylindrical coordinates are shown in Figure 22.1
below.
Figure 22.1
20 LAPLACE’S EQUATIONS IN CIRCULAR REGIONS 177
(a) Show that x = r cos θ, y = r sin θ, z = z.
(b) Show that
u
xx
+ u
yy
+ u
zz
= u
rr
+
1
r
u
r
+
1
r
2
u
θθ
+ u
zz
.
178SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
Sample Exam Questions
Exercise 20.11
An important result about harmonic functions is the so-called the maximum
principle which states: Any harmonic function u(x, y) deﬁned in a domain
Ω satisﬁes the inequality
min
(x,y)∈∂Ω
u(x, y) ≤ u(x, y) ≤ max
(x,y)∈∂Ω
u(x, y), ∀(x, y) ∈ Ω ∪ ∂Ω
where ∂Ω denotes the boundary of Ω.
Let u be harmonic in Ω = {(x, y) : x
2
+y
2
< 1} and satisﬁes u(x, y) = 2 −x
for all (x, y) ∈ ∂Ω. Show that u(x, y) > 0 for all (x, y) ∈ Ω.
Exercise 20.12
Let u be harmonic in Ω = {(x, y) : x
2
+y
2
< 1} and satisﬁes u(x, y) = 1+3x
for all (x, y) ∈ ∂Ω. Determine
(i) max
(x,y)∈Ω
u(x, y)
(ii) min
(x,y)∈Ω
u(x, y)
without solving ∆u = 0.
Exercise 20.13
Let u
1
(x, y) and u
2
(x, y) be harmonic functions on a smooth domain Ω such
that
u
1
|
∂Ω
= g
1
(x, y) and u
2
|
∂Ω
= g
3
(x, y)
where g
1
and g
2
are continuous functions satisfying
max
(x,y)∈∂Ω
g
1
(x, y) < min
(x,y)∈∂Ω
g
1
(x, y).
Prove that u
1
(x, y) < u
2
(x, y) for all (x, y) ∈ Ω ∪ ∂Ω.
Exercise 20.14
Show that r
n
cos (nθ) and r
n
sin (nθ) satisfy Laplace’s equation in polar co-
ordinates.
Exercise 20.15
Solve the Dirichlet problem
∆u = 0, 0 ≤ r < a, −π ≤ θ ≤ π
u(a, θ) = sin
2
θ.
20 LAPLACE’S EQUATIONS IN CIRCULAR REGIONS 179
Exercise 20.16
Solve Laplace’s equation
u
xx
+ u
yy
= 0
outside a circular disk (r ≥ a) subject to the boundary condition
u(a, θ) = ln 2 + 4 cos 3θ.
You may assume that the solution remains bounded as r →∞.
180SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
The Laplace Transform
Solutions for PDEs
If in a partial diﬀerential equation the time t is one of the independent vari-
ables of the searched-for function, we say that the PDE is an evolution
equation. Examples of evolutions equations are the heat equation and the
wave equation. In contrast, when the equation involves only spatial indepen-
dent variables then the equation is called a stationary equation. Examples
of stationary equations are the Laplace’s equations and Poisson equations.
There are classes of methods that can be used for solving the initial value or
initial boundary problems for evolution equations. We refer to these meth-
ods as the methods of integral transforms. The fundamental ones are the
Laplace and the Fourier transforms. In this chapter we will just consider the
Laplace transform.
21 Essentials of the Laplace Transform
Laplace transform has been introduced in an ODE course, and is used espe-
cially to solve linear ODEs with constant coeﬃcients, where the equations
are transformed to algebraic equations. This idea can be easily extended
to PDEs, where the transformation leads to the decrease of the number of
independent variables. PDEs in two variables are thus reduced to ODEs. In
this section we review the Laplace transform and its properties.
Laplace transform is yet another operational tool for solving constant coeﬃ-
cients linear diﬀerential equations. The process of solution consists of three
main steps:
• The given “hard” problem is transformed into a “simple” equation.
• This simple equation is solved by purely algebraic manipulations.
• The solution of the simple equation is transformed back to obtain the so-
181
182 THE LAPLACE TRANSFORM SOLUTIONS FOR PDES
lution of the given problem.
In this way the Laplace transformation reduces the problem of solving a dif-
ferential equation to an algebraic problem. The third step is made easier by
tables, whose role is similar to that of integral tables in integration.
The above procedure can be summarized by Figure 21.1
Figure 21.1
In this section we introduce the concept of Laplace transform and discuss
some of its properties.
The Laplace transform is deﬁned in the following way. Let f(t) be deﬁned
for t ≥ 0. Then the Laplace transform of f, which is denoted by L[f(t)]
or by F(s), is deﬁned by the following equation
L[f(t)] = F(s) = lim
T→∞
_
T
0
f(t)e
−st
dt =
_

0
f(t)e
−st
dt
The integral which deﬁnes a Laplace transform is an improper integral. An
improper integral may converge or diverge, depending on the integrand.
When the improper integral is convergent then we say that the function f(t)
possesses a Laplace transform. So what types of functions possess Laplace
transforms, that is, what type of functions guarantees a convergent improper
integral.
Example 21.1
Find the Laplace transform, if it exists, of each of the following functions
(a) f(t) = e
at
(b) f(t) = 1 (c) f(t) = t (d) f(t) = e
t
2
Solution.
(a) Using the deﬁnition of Laplace transform we see that
L[e
at
] =
_

0
e
−(s−a)t
dt = lim
T→∞
_
T
0
e
−(s−a)t
dt.
But
_
T
0
e
−(s−a)t
dt =
_
T if s = a
1−e
−(s−a)T
s−a
if s = a.
21 ESSENTIALS OF THE LAPLACE TRANSFORM 183
For the improper integral to converge we need s > a. In this case,
L[e
at
] = F(s) =
1
s −a
, s > a.
(b) In a similar way to what was done in part (a), we ﬁnd
L[1] =
_

0
e
−st
dt = lim
T→∞
_
T
0
e
−st
dt =
1
s
, s > 0.
(c) We have
L[t] =
_

0
te
−st
dt =
_

te
−st
s

e
−st
s
2
_

0
=
1
s
2
, s > 0.
(d) Again using the deﬁnition of Laplace transform we ﬁnd
L[e
t
2
] =
_

0
e
t
2
−st
dt.
If s ≤ 0 then t
2
−st ≥ 0 so that e
t
2
−st
≥ 1 and this implies that
_

0
e
t
2
−st
dt ≥
_

0
dt. Since the integral on the right is divergent, by the comparison theorem
of improper integrals (see Theorem 23.1 below) the integral on the left is also
divergent. Now, if s > 0 then
_

0
e
t(t−s)
dt ≥
_

s
dt. By the same reasoning
the integral on the left is divergent. This shows that the function f(t) = e
t
2
does not possess a Laplace transform
The above example raises the question of what class or classes of functions
possess a Laplace transform. To answer this question we introduce few math-
ematical concepts.
A function f that satisﬁes
|f(t)| ≤ Me
at
, t ≥ C (21.1)
is said to be a function with an exponential order at inﬁnity. A function
f is called piecewise continuous on an interval if the interval can be bro-
ken into a ﬁnite number of subintervals on which the function is continuous
on each open subinterval (i.e. the subinterval without its endpoints) and
has a ﬁnite limit at the endpoints (jump discontinuities and no vertical
asymptotes) of each subinterval. Below is a sketch of a piecewise continuous
184 THE LAPLACE TRANSFORM SOLUTIONS FOR PDES
function.
Note that a piecewise continuous function is a function that has a ﬁnite
number of breaks in it and doesn’t blow up to inﬁnity anywhere. A func-
tion deﬁned for t ≥ 0 is said to be piecewise continuous on the inﬁnite
interval if it is piecewise continuous on 0 ≤ t ≤ T for all T > 0.
Example 21.2
Show that the following functions are piecewise continuous and of exponential
order at inﬁnity for t ≥ 0
(a) f(t) = t
n
(b) f(t) = t
n
sin at
Solution.
(a) Since e
t
=

n=0
t
n
n!

t
n
n!
, t
n
≤ n!e
t
. Hence, t
n
is piecewise continuous and
of exponential order at inﬁnity.
(b) Since |t
n
sin at| ≤ n!e
t
, t
n
sin at is piecewise continuous and of exponential
order at inﬁnity
The following is an existence result of Laplace transform.
Theorem 21.1
Suppose that f(t) is piecewise continuous on t ≥ 0 and has an exponential
order at inﬁnity with |f(t)| ≤ Me
at
for t ≥ C. Then the Laplace transform
F(s) =
_

0
f(t)e
−st
dt
exists as long as s > a. Note that the two conditions above are suﬃcient, but
not necessary, for F(s) to exist.
In what follows, we will denote the class of all piecewise continuous functions
with exponential order at inﬁnity by PE. The next theorem shows that any
linear combination of functions in PE is also in PE. The same is true for the
product of two functions in PE.
21 ESSENTIALS OF THE LAPLACE TRANSFORM 185
Theorem 21.2
Suppose that f(t) and g(t) are two elements of PE with
|f(t)| ≤ M
1
e
a
1
t
, t ≥ C
1
and |g(t)| ≤ M
2
e
a
1
t
, t ≥ C
2
.
(i) For any constants α and β the function αf(t) +βg(t) is also a member of
PE. Moreover
L[αf(t) + βg(t)] = αL[f(t)] + βL[g(t)].
(ii) The function h(t) = f(t)g(t) is an element of PE.
We next discuss the problem of how to determine the function f(t) if F(s)
is given. That is, how do we invert the transform. The following result on
uniqueness provides a possible answer. This result establishes a one-to-one
correspondence between the set PE and its Laplace transforms. Alterna-
tively, the following theorem asserts that the Laplace transform of a member
in PE is unique.
Theorem 21.3
Let f(t) and g(t) be two elements in PE with Laplace transforms F(s) and
G(s) such that F(s) = G(s) for some s > a. Then f(t) = g(t) for all t ≥ 0
where both functions are continuous.
With the above theorem, we can now oﬃcially deﬁne the inverse Laplace
transform as follows: For a piecewise continuous function f of exponential
order at inﬁnity whose Laplace transform is F, we call f the inverse Laplace
transform of F and write f = L
−1
[F(s)]. Symbolically
f(t) = L
−1
[F(s)] ⇐⇒F(s) = L[f(t)].
Example 21.3
Find L
−1
_
1
s−1
_
, s > 1.
Solution.
From Example 23.1(a), we have that L[e
at
] =
1
s−a
, s > a. In particular, for
a = 1 we ﬁnd that L[e
t
] =
1
s−1
, s > 1. Hence, L
−1
_
1
s−1
_
= e
t
, t ≥ 0 .
The above theorem states that if f(t) is continuous and has a Laplace trans-
form F(s), then there is no other function that has the same Laplace trans-
form. To ﬁnd L
−1
[F(s)], we can inspect tables of Laplace transforms of
186 THE LAPLACE TRANSFORM SOLUTIONS FOR PDES
known functions to ﬁnd a particular f(t) that yields the given F(s).
When the function f(t) is not continuous, the uniqueness of the inverse
Laplace transform is not assured. The following example addresses the
uniqueness issue.
Example 21.4
Consider the two functions f(t) = H(t)H(3 −t) and g(t) = H(t) −H(t −3),
where H is the Heaviside function deﬁned by
H(t) =
_
1, t ≥ 0
0, t < 0
(a) Are the two functions identical?
(b) Show that L[f(t)] = L[g(t).
Solution.
(a) We have
f(t) =
_
1, 0 ≤ t ≤ 3
0, t > 3
and
g(t) =
_
1, 0 ≤ t < 3
0, t ≥ 3
Since f(3) = 1 and g(3) = 0 then f and g are not identical.
(b) We have
L[f(t)] = L[g(t)] =
_
3
0
e
−st
dt =
1 −e
−3s
s
, s > 0.
Thus, both functions f(t) and g(t) have the same Laplace transform even
though they are not identical. However, they are equal on the interval(s)
where they are both continuous
The inverse Laplace transform possesses a linear property as indicated in
the following result.
Theorem 21.4
Given two Laplace transforms F(s) and G(s) then
L
−1
[aF(s) + bG(s)] = aL
−1
[F(s)] + bL
−1
[G(s)]
for any constants a and b.
21 ESSENTIALS OF THE LAPLACE TRANSFORM 187
Convolution integrals are useful when ﬁnding the inverse Laplace transform
of products. They are deﬁned as follows: The convolution of two scalar
piecewise continuous functions f(t) and g(t) deﬁned for t ≥ 0 is the integral
(f ∗ g)(t) =
_
t
0
f(t −s)g(s)ds.
Example 21.5
Find f ∗ g where f(t) = e
−t
and g(t) = sin t.
Solution.
Using integration by parts twice we arrive at
(f ∗ g)(t) =
_
t
0
e
−(t−s)
sin sds
=
1
2
_
e
−(t−s)
(sin s −cos s)
¸
t
0
=
e
−t
2
+
1
2
(sin t −cos t)
Next, we state several properties of convolution product, which resemble
those of ordinary product.
Theorem 21.5
Let f(t), g(t), and k(t) be three piecewise continuous scalar functions deﬁned
for t ≥ 0 and c
1
and c
2
are arbitrary constants. Then
(i) f ∗ g = g ∗ f (Commutative Law)
(ii) (f ∗ g) ∗ k = f ∗ (g ∗ k) (Associative Law)
(iii) f ∗ (c
1
g + c
2
k) = c
1
f ∗ g + c
2
f ∗ k (Distributive Law)
Example 21.6
Express the solution to the initial value problem y

+ αy = g(t), y(0) = y
0
in terms of a convolution integral.
Solution.
Solving this initial value problem by the method of integrating factor we ﬁnd
y(t) = e
−αt
y
0
+
_
t
0
e
−α(t−s)
g(s)ds = e
−αt
y
0
+ (e
−αt
∗ g)(t)
The following theorem, known as the Convolution Theorem, provides a way
for ﬁnding the Laplace transform of a convolution integral and also ﬁnding
the inverse Laplace transform of a product.
188 THE LAPLACE TRANSFORM SOLUTIONS FOR PDES
Theorem 21.6
If f(t) and g(t) are piecewise continuous for t ≥ 0, and of exponential order
at inﬁnity then
L[(f ∗ g)(t)] = L[f(t)]L[g(t)] = F(s)G(s).
Thus, (f ∗ g)(t) = L
−1
[F(s)G(s)].
Example 21.7
Use the convolution theorem to ﬁnd the inverse Laplace transform of
P(s) =
1
(s
2
+ a
2
)
2
.
Solution.
Note that
P(s) =
_
1
s
2
+ a
2
__
1
s
2
+ a
2
_
.
So, in this case we have, F(s) = G(s) =
1
s
2
+a
2
so that f(t) = g(t) =
1
a
sin (at).
Thus,
(f ∗ g)(t) =
1
a
2
_
t
0
sin (at −as) sin (as)ds =
1
2a
3
(sin (at) −at cos (at))
Example 21.8
Solve the initial value problem
4y

+ y = g(t), y(0) = 3, y

(0) = −7
Solution.
Take the Laplace transform of all the terms and plug in the initial conditions
to obtain
4(s
2
Y (s) −3s + 7) + Y (s) = G(s)
or
(4s
2
+ 1)Y (s) −12s + 28 = G(s).
Solving for Y (s) we ﬁnd
21 ESSENTIALS OF THE LAPLACE TRANSFORM 189
Y (s) =
12s −28
4
_
s
2
+
1
4
_ +
G(s)
4
_
s
2
+
1
4
_
=
3s
s
2
+
_
(
1
2
_
2
−7
_
1
2
_
2
s
2
+
_
1
2
_
2
+
1
4
G(s)
_
1
2
_
2
s
2
+
_
1
2
_
2
Hence,
y(t) = 3 cos
_
t
2
_
−7 sin
_
t
2
_
+
1
2
_
t
0
sin
_
s
2
_
g(t −s)ds.
So, once we decide on a g(t) all we need to do is to evaluate the integral and
we’ll have the solution
We conclude this section with the following table of Laplace transform pairs.
H is the Heaviside function and δ is the Dirac delta function as deﬁned in
Section 0.
190 THE LAPLACE TRANSFORM SOLUTIONS FOR PDES
f(t) F(s)
H(t) =
_
1, t ≥ 0
0, t < 0
1
s
, s > 0
t
n
, n = 1, 2, · · ·
n!
s
n+1
, s > 0
e
αt s
s−α
, s > α
sin (ωt)
ω
s
2

2
, s > 0
cos (ωt)
s
s
2

2
, s > 0
sinh (ωt)
ω
s
2
−ω
2
, s > |ω|
cosh (ωt)
s
s
2
−ω
2
, s > |ω|
e
αt
f(t), with |f(t)| ≤ Me
at
F(s −α), s > α +a
e
αt
H(t)
1
s−α
, s > α
e
αt
t
n
, n = 1, 2, · · ·
n!
(s−α)
n+1
, s > α
e
αt
sin (ωt)
ω
(s−α)
2

2
, s > α
e
αt
cos (ωt)
s−α
(s−α)
2

2
, s > α
f(t −α)H(t −α), α ≥ 0 e
−αs
F(s), s > a
with |f(t)| ≤ Me
at
H(t −α), α ≥ 0
e
−αs
s
, s > 0
tf(t) -F

(s)
t

sin ωt
s
(s
2

2
)
2
, s > 0
1

3
[sin ωt −ωt cos ωt]
1
(s
2

2
)
2
, s > 0
f

(t), with f(t) continuous sF(s) −f(0)
and |f

(t)| ≤ Me
at
s > max{a, 0} + 1
f

(t), with f

(t) continuous s
2
F(s) −sf(0) −f

(0)
and |f

(t)| ≤ Me
at
s > max{a, 0} + 1
f
(n)
(t), with f
(n−1)
(t) continuous s
n
F(s) −s
n−1
f(0) −· · ·
and |f
(n)
(t)| ≤ Me
at
-sf
(n−2)
(0) −f
(n−1)
(0)
s > max{a, 0} + 1
_
t
0
f(u)du, with |f(t)| ≤ Me
at
F(s)
s
, s > max{a, 0} + 1
Table L
21 ESSENTIALS OF THE LAPLACE TRANSFORM 191
Practice Problems
Exercise 21.1
Determine whether the integral
_

0
1
1+t
2
dt converges. If the integral con-
verges, give its value.
Exercise 21.2
Determine whether the integral
_

0
t
1+t
2
dt converges. If the integral con-
verges, give its value.
Exercise 21.3
Determine whether the integral
_

0
e
−t
cos (e
−t
)dt converges. If the integral
converges, give its value.
Exercise 21.4
Using the deﬁnition, ﬁnd L[e
3t
], if it exists. If the Laplace transform exists
then ﬁnd the domain of F(s).
Exercise 21.5
Using the deﬁnition, ﬁnd L[t −5], if it exists. If the Laplace transform exists
then ﬁnd the domain of F(s).
Exercise 21.6
Using the deﬁnition, ﬁnd L[e
(t−1)
2
], if it exists. If the Laplace transform
exists then ﬁnd the domain of F(s).
Exercise 21.7
Using the deﬁnition, ﬁnd L[(t − 2)
2
], if it exists. If the Laplace transform
exists then ﬁnd the domain of F(s).
Exercise 21.8
Using the deﬁnition, ﬁnd L[f(t)], if it exists. If the Laplace transform exists
then ﬁnd the domain of F(s).
f(t) =
_
0, 0 ≤ t < 1
t −1, t ≥ 1
192 THE LAPLACE TRANSFORM SOLUTIONS FOR PDES
Exercise 21.9
Using the deﬁnition, ﬁnd L[f(t)], if it exists. If the Laplace transform exists
then ﬁnd the domain of F(s).
f(t) =
_
_
_
0, 0 ≤ t < 1
t −1, 1 ≤ t < 2
0, t ≥ 2.
Exercise 21.10
Let n be a positive integer. Using integration by parts establish the reduction
formula
_
t
n
e
−st
dt = −
t
n
e
−st
s
+
n
s
_
t
n−1
e
−st
dt, s > 0.
Exercise 21.11
For s > 0 and n a positive integer evaluate the limits
(a) lim
t→0
t
n
e
−st
(b) lim
t→∞
t
n
e
−st
Exercise 21.12
Use the linearity property of Laplace transform to ﬁnd L[5e
−7t
+ t + 2e
2t
].
Find the domain of F(s).
Exercise 21.13
Find L
−1
_
3
s−2
_
.
Exercise 21.14
Find L
−1
_

2
s
2
+
1
s+1
_
.
Exercise 21.15
Find L
−1
_
2
s+2
+
2
s−2
_
.
Exercise 21.16
Use Table L to ﬁnd L[2e
t
+ 5].
Exercise 21.17
Use Table L to ﬁnd L[e
3t−3
H(t −1)].
Exercise 21.18
Use Table L to ﬁnd L[sin
2
ωt].
21 ESSENTIALS OF THE LAPLACE TRANSFORM 193
Exercise 21.19
Use Table L to ﬁnd L[sin 3t cos 3t].
Exercise 21.20
Use Table L to ﬁnd L[e
2t
cos 3t].
Exercise 21.21
Use Table L to ﬁnd L[e
4t
(t
2
+ 3t + 5)].
Exercise 21.22
Use Table L to ﬁnd L
−1
[
10
s
2
+25
+
4
s−3
].
Exercise 21.23
Use Table L to ﬁnd L
−1
[
5
(s−3)
4
].
194 THE LAPLACE TRANSFORM SOLUTIONS FOR PDES
Sample Exam Questions
Exercise 21.24
Use Table L to ﬁnd L
−1
[
e
−2s
s−9
].
Exercise 21.25
Using the partial fraction decomposition ﬁnd L
−1
_
12
(s−3)(s+1)
_
.
Exercise 21.26
Using the partial fraction decomposition ﬁnd L
−1
_
24e
−5s
s
2
−9
_
.
Exercise 21.27
Use Laplace transform technique to solve the initial value problem
y

+ 4y = g(t), y(0) = 2
where
g(t) =
_
_
_
0, 0 ≤ t < 1
12, 1 ≤ t < 3
0, t ≥ 3
Exercise 21.28
Use Laplace transform technique to solve the initial value problem
y

−4y = e
3t
, y(0) = 0, y

(0) = 0.
Exercise 21.29
Consider the functions f(t) = e
t
and g(t) = e
−2t
, t ≥ 0. Compute f ∗ g in
two diﬀerent ways.
(a) By directly evaluating the integral.
(b) By computing L
−1
[F(s)G(s)] where F(s) = L[f(t)] and G(s) = L[g(t)].
Exercise 21.30
Consider the functions f(t) = sin t and g(t) = cos t, t ≥ 0. Compute f ∗ g in
two diﬀerent ways.
(a) By directly evaluating the integral.
(b) By computing L
−1
[F(s)G(s)] where F(s) = L[f(t)] and G(s) = L[g(t)].
Exercise 21.31
Compute t ∗ t ∗ t.
21 ESSENTIALS OF THE LAPLACE TRANSFORM 195
Exercise 21.32
Compute H(t) ∗ e
−t
∗ e
−2t
.
Exercise 21.33
Compute t ∗ e
−t
∗ e
t
.
196 THE LAPLACE TRANSFORM SOLUTIONS FOR PDES
22 Solving PDEs Using Laplace Transform
The same idea for solving linear ODEs using Laplace transform can be ex-
ploited when solving PDEs for functions in two variables u = u(x, t). The
transformation will be done with respect to the time variable t ≥ 0, the spa-
tial variable x will be treated as a parameter unaﬀected by this transform.
In particular we deﬁne the Laplace transform of u(x, t) by the formula
L(u(x, t)) = U(x, s) =
_

0
u(x, τ)e
−sτ
dτ.
The time derivatives are transformed in the same way as in the case of
functions in one variable, that is, for example
L(u
t
)(x, t) = sU(x, s) −u(x, 0)
and
L(u
tt
)(x, s) = s
2
U(x, s) −su(x, 0) −u
t
(x, 0).
The spatial derivatives remain unchanged, for example,
Lu
x
(x, t) =
_

0
u
x
(x, τ)e
−sτ
dτ =

∂x
_

0
u(x, τ)e
−sτ
dτ = U
x
(x, s).
Likewise, we have
Lu
xx
(x, t) = U
xx
(x, s).
Thus, applying the Laplace transform to a PDE in two variables x and t we
obtain an ODE in the variable x and with the parameter s.
Example 22.1
Let u(x, t) be the concentration of a chemical contaminant dissolved in a
liquid on a half-inﬁnte domain x > 0. Let us assume that at time t = 0 the
concentration is 0 and on the boundary x = 0, constant unit concentration of
the contaminant is kept for t > 0. The behaviour of this problem is described
by the following mathematical model
_
¸
¸
_
¸
¸
_
u
t
−u
xx
= 0 , x > 0, t > 0
u(x, 0) = 0,
u(0, t) = 1,
|u(x, t)| < ∞
Find u(x, t).
22 SOLVING PDES USING LAPLACE TRANSFORM 197
Solution.
Applying Laplace transform to both sides of the equation we obtain
sU(x, s) −u(x, 0) −U
xx
(x, s) = 0
or
U
xx
(x, s) −sU(x, s) = 0.
This is a second order linear ODE in the variable x and positive parameter
s. Its general solution is
U(x, s) = A(s)e

sx
+ B(s)e

sx
.
Since U(x, s) is bounded in both variables, we must have B(s) = 0 and in
this case we obtain
U(x, s) = A(s)e

sx
.
Next, we apply Laplace transform to the boundary condition obtaining
U(0, s) = L(1) =
1
s
.
This leads to A(s) =
1
s
and the transformed solution becomes
U(x, s) =
1
s
e

sx
.
Thus,
u(x, t) = L
−1
_
1
s
e

sx
_
.
One can use a software package to ﬁnd the expression for L
−1
_
1
s
e

sx
_
Example 22.2
Solve the following initial boundary value problem
_
¸
¸
_
¸
¸
_
u
t
−u
xx
= 0 , x > 0, t > 0
u(x, 0) = 0,
u(0, t) = f(t),
|u(x, t)| < ∞
198 THE LAPLACE TRANSFORM SOLUTIONS FOR PDES
Solution.
Following the argument of the previous example we ﬁnd
U(x, s) = F(s)e

sx
, F(s) = Lf(t).
Thus, using Theorem 21.6 we can write
u(x, t) = L
−1
_
F(s)e

sx
_
= f ∗ L
−1
(e

sx
).
It can be shown that
L
−1
(e

sx
) =
x

4πt
3
e

x
2
4t
.
Hence,
u(x, t) =
_
t
0
x
_
4π(t −s)
3
e

x
2
4(t−s)
f(s)ds
Example 22.3
Solve the wave equation
_
¸
¸
_
¸
¸
_
u
tt
−c
2
u
xx
= 0 , x > 0, t > 0
u(x, 0) = u
t
(x, 0) = 0,
u(0, t) = f(t),
|u(x, t)| < ∞
Solution.
Applying Laplace transform to both sides of the equation we obtain
s
2
U(x, s) −su(x, 0) −u
t
(x, 0) −c
2
U
xx
(x, s) = 0
or
c
2
U
xx
(x, s) −s
2
U(x, s) = 0.
This is a second order linear ODE in the variable x and positive parameter
s. Its general solution is
U(x, s) = A(s)e

s
c
x
+ B(s)e
s
c
x
.
Since U(x, s) is bounded, we must have B(s) = 0 and in this case we obtain
U(x, s) = A(s)e

s
c
x
.
22 SOLVING PDES USING LAPLACE TRANSFORM 199
Next, we apply Laplace transform to the boundary condition obtaining
U(0, s) = L(f(t)) = F(s).
This leads to A(s) = F(s) and the transformed solution becomes
U(x, s) = F(s)e

s
c
x
.
Thus,
u(x, t) = L
−1
_
F(s)e

x
c
s
_
= H
_
t −
x
c
_
f
_
t −
x
c
_
Remark 22.1
Laplace transforms are useful in solving parabolic and some hyperbolic PDEs.
They are not in general useful in solving elliptic PDEs.
200 THE LAPLACE TRANSFORM SOLUTIONS FOR PDES
Practice Problems
Exercise 22.1
Solve by Laplace transform
_
_
_
u
t
+ u
x
= 0 , x > 0, t > 0
u(x, 0) = sin x,
u(0, t) = 0
Hint: Method of integrating factor of ODEs.
Exercise 22.2
Solve by Laplace transform
_
_
_
u
t
+ u
x
= −u , x > 0, t > 0
u(x, 0) = sin x,
u(0, t) = 0
Exercise 22.3
Solve
u
t
= 4u
xx
u(0, t) = u(1, t) = 0
u(x, 0) = 2 sin πx + 3 sin 2πx.
Hint: A particular solution of a second order ODE must be found using the
method of variation of parameters.
Exercise 22.4
Solve by Laplace transform
_
_
_
u
t
−u
x
= u , x > 0, t > 0
u(x, 0) = e
−5x
,
|u(x, t)| < ∞
Exercise 22.5
Solve by Laplace transform
_
_
_
u
t
+ u
x
= t , x > 0, t > 0
u(x, 0) = 0,
u(0, t) = t
2
22 SOLVING PDES USING LAPLACE TRANSFORM 201
Exercise 22.6
Solve by Laplace transform
_
_
_
xu
t
+ u
x
= 0 , x > 0, t > 0
u(x, 0) = 0,
u(0, t) = t
Exercise 22.7
Solve by Laplace transform
_
¸
¸
_
¸
¸
_
u
tt
−c
2
u
xx
= 0 , x > 0, t > 0
u(x, 0) = u
t
(x, 0) = 0,
u(0, t) = sin x,
|u(x, t)| < ∞
Exercise 22.8
Solve by Laplace transform
u
tt
−9u
xx
= 0, 0 ≤ x ≤ π, t > 0
u(0, t) = u(π, t) = 0,
u
t
(x, 0) = 0, u(x, 0) = 2 sin x.
Exercise 22.9
Solve by Laplace transform
_
_
_
u
xy
= 1 , x > 0, y > 0
u(x, 0) = 1,
u(0, y) = y + 1.
Exercise 22.10
Solve by Laplace transform
_
¸
¸
_
¸
¸
_
u
tt
= c
2
u
xx
, x > 0, t > 0
u(x, 0) = u
t
(x, 0) = 0,
u
x
(0, t) = f(t),
|u(x, t)| < ∞.
202 THE LAPLACE TRANSFORM SOLUTIONS FOR PDES
Sample Exam Questions
Exercise 22.11
Solve by Laplace transform
_
_
_
u
t
+ u
x
= u , x > 0, t > 0
u(x, 0) = sin x,
u(0, t) = 0
Exercise 22.12
Solve by Laplace transform
_
¸
¸
_
¸
¸
_
u
t
−c
2
u
xx
= 0 , x > 0, t > 0
u(x, 0) = T,
u(0, t) = 0,
|u(x, t)| < ∞
Exercise 22.13
Solve by Laplace transform
u
t
−3u
xx
= 0, 0 ≤ x ≤ 2, t > 0
u(0, t) = u(2, t) = 0,
u(x, 0) = 5 sin (πx)
Exercise 22.14
Solve by Laplace transform
u
t
−4u
xx
= 0, 0 ≤ x ≤ π, t > 0
u
x
(0, t) = u(π, t) = 0,
u(x, 0) = 40 cos
x
2
Exercise 22.15
Solve by Laplace transform
u
tt
−4u
xx
= 0, 0 ≤ x ≤ 2, t > 0
u(0, t) = u(2, t) = 0,
u
t
(x, 0) = 0, u(x, 0) = 3 sin πx.
The Fourier Transform
Solutions for PDEs
In the previous chapter we discussed one class of integral transform meth-
ods, the Laplace transfom. In this chapter, we consider a second fundamental
class of integral transform methods, the so-called Fourier transform.
Fourier series are designed to solve boundary value problems on bounded
intervals. The extension of Fourier methods to the entire real line leads nat-
urally to the Fourier transform, an extremely powerful mathematical tool for
the analysis of non-periodic functions. The Fourier transform is of fundamen-
tal importance in a broad range of applications, including both ordinary and
partial diﬀerential equations, quantum mechanics, signal processing, control
theory, and probability, to name but a few.
23 Complex Version of Fourier Series
We have seen in Section 15 that a 2L−periodic function f : R → R that is
piecewise smooth on [−L, L] can be expanded in a Fourier series
f(x) =
a
0
2
+

n=1
_
a
n
cos
_

L
x
_
+ b
n
sin
_

L
x
__
at all points of continuity of f. In the context of Fourier analysis, this is
referred to as the real form of the Fourier series. It is often convenient to
recast this series in complex form by means of Euler formula
e
ix
= cos x + i sin x.
It follows from this formula that
203
204 THE FOURIER TRANSFORM SOLUTIONS FOR PDES
e
ix
+ e
−ix
= 2 cos x and e
ix
−e
−ix
= 2i sin x
or
cos x =
e
ix
+e
−ix
2
and sin x =
e
ix
−e
−ix
2i
.
Hence the Fourier expansion of f can be rewritten as
f(x) =
a
0
2
+

n=1
_
a
n
_
e
inπx
L
+ e

inπx
L
2
_
+b
n
_
e
inπx
L
−e

inπx
L
2i
__
f(x) =

−∞
c
n
e
inπx
L
(23.1)
where c
0
=
a
0
2
and for n ∈ N we have
c
n
=
a
n
−ib
n
2
c
−n
=
a
n
+ ib
n
2
.
It follows that if n ∈ N then
a
n
= c
n
+ c
−n
and b
n
= i(c
n
−c
−n
). (23.2)
That is, a
n
and b
n
can be easily found once we have formulas for c
n
. In order
to ﬁnd these formulas, we need to evaluate the following integral
_
L
−L
e
inπx
L
e

imπx
L
dx =
_
L
−L
e
i(n−m)πx
L
dx
=
L
i(n −m)π
e
i(n−m)πx
L
_
L
−L
=−
iL
(n −m)π
[cos [(n −m)π] + i sin [(n −m)π]
−cos [−(n −m)π] −i sin [−(n −m)π]]
=0
23 COMPLEX VERSION OF FOURIER SERIES 205
if n = m. If n = m then
_
L
−L
e
inπx
L
e

inπx
L
dx = 2L.
Now, if we multiply (23.1) by e

inπx
L
and integrate from −L to L and apply
the last result we ﬁnd
_
L
−L
f(x)e

inπx
L
dx = 2Lc
n
which yields the formula for coeﬃcients of the complex form of the Fourier
series:
c
n
=
1
2L
_
L
−L
f(x)e

inπx
L
dx, n = 0, ±1, ±2, · · · .
Example 23.1
Find the complex Fourier coeﬃcients of the function
f(x) = x, −π ≤ x ≤ π
extended to be periodic of period 2π.
Solution.
Using integration by parts and the fact that e

= e
−iπ
= −1 we ﬁnd
c
n
=
1

_
π
−π
xe
−inx
dx
=
1

__
ix
n
_
e
−inx
¸
¸
¸
¸
π
−π

_
π
−π
_
i
n
_
e
−inx
dx
_
=
1

__

n
_
e
−inπ
+
_

n
_
e
inπ
_
+
1

_
1
n
2
e
−inπ

1
n
2
e
inπ
_
=
1

_
2i
π
n
(−1)
n
_
+
1

(0) =
(−1)
n
i
n
Remark 23.1
It is often the case that the complex form of the Fourier series is far simpler
to calculate than the real form. One can then use (23.2) to ﬁnd the real form
of the Fourier series. For example, the Fourier coeﬃcients of the real form of
the previous function are given by
a
n
= (c
n
+ c
−n
) = 0 and b
n
= i(c
n
−c
−n
) =
2
n
(−1)
n+1
, n ∈ N
206 THE FOURIER TRANSFORM SOLUTIONS FOR PDES
Practice Problems
Exercise 23.1
Find the complex Fourier coeﬃcients of the function
f(x) = x, −1 ≤ x ≤ 1
extended to be periodic of period 2.
Exercise 23.2
Let
f(x) =
_
_
_
0 −π < x <
−π
2
1
−π
2
< x <
π
2
0 π < x < π
be 2π−periodic. Find its complex series representation.
Exercise 23.3
Find the complex Fourier series of the 2π−periodic function f(x) = e
ax
over
the interval (−π, π).
Exercise 23.4
Find the complex Fourier series of the 2π−periodic function f(x) = sin x
over the interval (−π, π).
Exercise 23.5
Find the complex Fourier series of the 2π−periodic function deﬁned
f(x) =
_
1 0 < x < T
0 T < x < 2π
Exercise 23.6
Let f(x) = x
2
, −π < x < π, be 2π−periodic.
(a) Calculate the complex Fourier series representation of f.
(b) Using the complex Fourier series found in (a), recover the real Fourier
series representation of f.
Exercise 23.7
Let f(x) = sin nπx, −
1
2
< x <
1
2
, be of period 1.
(a) Calculate the coeﬃcients a
n
, b
n
and c
n
.
(b) Find the complex Fourier series representation of f.
23 COMPLEX VERSION OF FOURIER SERIES 207
Exercise 23.8
Let f(x) = 2 −x, −2 < x < 2, be of period 2.
(a) Calculate the coeﬃcients a
n
, b
n
and c
n
.
(b) Find the complex Fourier series representation of f.
Exercise 23.9
Suppose that the coeﬃcients c
n
of the complex Fourier series are given by
c
n
=
_
2
iπn
if |n| is odd
0 if |n| is even.
Find a
n
, n = 0, 1, 2, · · · and b
n
, n = 1, 2, · · · .
Exercise 23.10
Recall that any complex number z can be written as z = Re(z) + iIm(z)
where Re(z) is called the real part of z and Im(z) is called the imaginary
part. The complex conjugate of z is the complex number z = Re(z) −
iIm(z). Using these deﬁnitions show that a
n
= 2Re(c
n
) and b
n
= −2Im(c
n
).
208 THE FOURIER TRANSFORM SOLUTIONS FOR PDES
Sample Exam Questions
Exercise 23.11
Suppose that
c
n
=
_
i
2πn
[e
−inT
−1] if n = 0
T

if n = 0.
Find a
n
and b
n
.
Exercise 23.12
Find the complex Fourier series of the function f(x) = e
x
on [−2, 2].
Exercise 23.13
Consider the wave form
(a) Write f(x) explicitly. What is the period of f.
(b) Determine a
0
and a
n
for n ∈ N.
(c) Determine b
n
for n ∈ N.
(d) Determine c
0
and c
n
for n ∈ N.
Exercise 23.14
If z is a complex number we deﬁne sin z =
1
2
(e
iz
− e
−iz
). Find the complex
form of the Fourier series for sin 3x without evaluating any integrals.
Exercise 23.15
Find c
n
for the 2π−periodic function
f(x) =
_
1 if s ≤ x ≤ s + h
0 elsewhere in [−π, π]
24 TWO DIMENSIONAL FOURIER TRANSFORMS 209
24 Two Dimensional Fourier Transforms
One of the problems with the theory of Fourier series discussed so far is that
it applies only to periodic functions. There are many times when one would
like to divide a function which is not periodic into a superposition of sines
and cosines. The Fourier transform is the tool often used for this purpose.
Like the Laplace transform, the Fourier transform is often an eﬀective tool
in ﬁnding explicit solutions to diﬀerential equations.
To start with, let f : R →R be a piecewise continuous function that vanishes
outside an interval of the form [−πL, πL]. This function can be extended to
a periodic function, still denoted by f, of period 2πL. From the previous
section we can ﬁnd the complex Fourier series of f to be
f(x) =

n=−∞
c
n
e
inx
L
(24.1)
where
c
n
=
1
2πL
_
πL
−πL
f(x)e

inx
L
.
Let ξ ∈ R. Multiply both sides of (24.1) by e
−iξx
and then integrate both sides
from −πL to πL. Assuming integration and summation can be interchanged
we ﬁnd
_
πL
−πL
f(x)e
−iξx
dx =

n=−∞
c
n
_
πL
−πL
e
−iξx
e
inx
L
dx.
It can be shown that the RHS converges, say to
ˆ
f(ξ), as L →∞. Hence, we
ﬁnd
ˆ
f(ξ) =
_

−∞
f(x)e
−iξx
dx (24.2)
The function
ˆ
f is called the Fourier transformof f. We will use the notation
F[f(x)] =
ˆ
f(ξ).
Now, letting ξ =
n
L
we ﬁnd
ˆ
f
_
n
L
_
=
_
πL
−πL
f(x)e

inx
L
dx = 2πLc
n
.
210 THE FOURIER TRANSFORM SOLUTIONS FOR PDES
Hence, (24.1) can be written in the form
f(x) =
1
2πL

n=−∞
ˆ
f
_
n
L
_
e
inx
L
.
In the limit as L → ∞, it can be shown that this last sum approaches an
improper integral, and our formula becomes
F
−1
[
ˆ
f(ξ)] = f(x) =
1

_

−∞
ˆ
f(ξ)e
iξx
dξ (24.3)
Equation (24.3) is called the Fourier inversion formula. If we make use of
Euler’s formula, we can write the Fourier inversion formula in terms of sines
and cosines,
f(x) =
1

_

−∞
ˆ
f(ξ) cos ξxdξ +
i

_

−∞
ˆ
f(ξ) sin ξxdξ
a superposition of sines and cosines of various frequencies.
Equations (24.2) and (24.3) allow one to pass back and forth between a given
function and its representation as a superposition of oscillations of various
frequencies.
Example 24.1
Find the Fourier transform of the function f(x) deﬁned by
f(x) =
_
e
−ax
if x ≥ 0
0 if x < 0
for some a > 0.
Solution.
We have
ˆ
f(ξ) =
_

−∞
f(t)e
−iξx
dx =
_

0
e
−ax
e
−iξx
dx
=
_

0
e
−ax−iξx
dx =
e
−x(a+iξ)
−(a + iξ)
¸
¸
¸
¸

0
=
1
a + iξ
The following theorem lists the basic properties of Fourier transform
24 TWO DIMENSIONAL FOURIER TRANSFORMS 211
Theorem 24.1
Let f, g, be piecewise continuous functions. Then we have the following
properties:
(1) Linearity: F[αf(x) +βg(x)] = αF[f(x)] +βF[g(x)], where α and β are
arbitrary numbers.
(2) Shifting: F[f(x −α)] = e
−αξ
F[f(x)].
(3) Scaling: F[f
_
x
α
_
] = αF[f(αx)].
(4) Continuity: If
_

−∞
|f(x)|dx < ∞ then
ˆ
f is continuous in ξ.
(5) Diﬀerentiation: F[f
(n)
(x)] = (iξ)
n
F[f(x)).
(6) Integration: F
__
x
0
f(s)ds
¸
= −
1

F[f(x)].
(7) Parseval’s Relation:
_

−∞
|f(x)|
2
dx =
1

_

−∞
|
ˆ
f(ξ)|
2
dξ.
(8) Duality: F[F[f(x)]] = 2πf(−x).
(9) Multiplication by x
n
: F[x
n
f(x)] = i
n
ˆ
f
(n)
(ξ).
(10) Gaussians: F[e
−αx
2
] =
_
π
α
e

ξ
2

.
(11) Product: F[(f(x)g(x)] =
1

F[f(x)] ∗ F[g(x)].
(12) Convolution: F[(f ∗ g)(x)] = F[f(x)] · F[g(x)].
Example 24.2
Determine the Fourier transform of the Gaussian u(x) = e
−αx
2
, α > 0.
Solution.
We have
ˆ u(ξ) =
_

−∞
e
−αx
2
e
−iξx
dx.
If we diﬀerentiate this relation with respect to the variable ξ and then inte-
grate by parts we obtain
ˆ u

(ξ) =−i
_

−∞
xe
−αx
2
e
−iξx
dx
=
i

_

−∞
d
dx
(e
−αx
2
)e
−iξx
dx
=

_

−∞
(e
−αx
2
)e
−iξx
dx = −
ξ

ˆ u(ξ)
Thus we have arrived at the ODE ˆ u

(ξ) = −
ξ

ˆ y(ξ) whose general solution
has the form
ˆ u(ξ) = Ce

ξ
2

212 THE FOURIER TRANSFORM SOLUTIONS FOR PDES
Since
ˆ u(0) =
_

−∞
e
−αx
dx =
_
π
α
= C
we ﬁnd
ˆ u(ξ) =
_
π
α
e

ξ
2

Example 24.3
Prove
F[f(−x)] =
ˆ
f(−ξ).
Solution.
Using a change of variables we ﬁnd
F[f(−x)] =
_

−∞
f(−x)e
−iξx
dx =
_

−∞
f(x)e
iξx
dx =
ˆ
f(−ξ)
Example 24.4
Prove
F[F[f(x)]] = 2πf(−x).
Solution.
We have
f(x) =
1

_

−∞
ˆ
f(ξ)e
iξx

Thus,
2πf(−x) =
_

−∞
ˆ
f(ξ)e
−iξx
dξ = F[
ˆ
f(ξ)] = F[F[f(x)]]
The following theorem lists the properties of inverse Fourier transform
Theorem 24.2
Let f and g be piecewise continuous functions.
(1’) Linearity: F
−1

ˆ
f(ξ) + βˆ g(ξ)] = αF
−1
[
ˆ
f(ξ)] + βF
−1
[ˆ g(ξ)].
(2’) Derivatives: F
−1
[
ˆ
f
(n)
(ξ)] = (−ix)
n
f(x).
(3’) Multiplication by ξ
n
: F
−1

n
ˆ
f(ξ)] = (−i)
n
f
(n)
(x).
(4’) Multiplication by e
−iξα
: F
−1
[e
−iξα
ˆ
f(ξ)] = f(x −α).
(5’) Gaussians: F
−1
[e
−αξ
2
] =
1

4πα
e

x
2

.
(6’) Product: F
−1
[
ˆ
f(ξ)ˆ g(ξ)] = f(x) ∗ g(x).
(7’) Convolution: F
−1
[
ˆ
f ∗ ˆ g(ξ)] = 2π(fg)(x).
24 TWO DIMENSIONAL FOURIER TRANSFORMS 213
Remark 24.1
It is important to mention that there exists no established convention how
to deﬁne the Fourier transform. In literature, we can meet an equivalent
deﬁnition of (24.2) with the constant
1

or
1

in front of the integral.
There also exist deﬁnitions with positive sign in the exponent. The reader
should keep this fact in mind while working with various sources or using the
transformation tables.
214 THE FOURIER TRANSFORM SOLUTIONS FOR PDES
Practice Problems
Exercise 24.1
Find the Fourier transform of the function
f(x) =
_
1 if −1 ≤ x ≤ 1
0 otherwise.
Exercise 24.2
Obtain the transformed problem when applying the Fourier transform with
respect to the spatial variable to the equation and initial condition
u
t
+ cu
x
= 0
u(x, 0) = f(x).
Exercise 24.3
Obtain the transformed problem when applying the Fourier transform with
respect to the spatial variable to the equation and both initial conditions
u
tt
= c
2
u
xx
, x ∈ R, t > 0
u(x, 0) = f(x)
u
t
(x, 0) = g(x).
Exercise 24.4
Obtain the transformed problem when applying the Fourier transform with
respect to the spatial variable to the equation and both initial conditions
∆u = u
xx
+ u
yy
= 0, x ∈ R, 0 < y < L
u(x, 0) = 0
u(x, L) =
_
1 if −a < x < a
0 otherwise
Exercise 24.5
Find the Fourier transform of f(x) = e
−|x|α
, where α > 0.
24 TWO DIMENSIONAL FOURIER TRANSFORMS 215
Exercise 24.6
Prove that
F[e
−x
H(x)] =
1
1 + iξ
where
H(x) =
_
1 if x ≥ 0
0 otherwise.
Exercise 24.7
Prove that
F
_
1
1 + ix
_
= 2πe
ξ
H(−ξ).
Exercise 24.8
Prove
F[f(x −α)] = e
−iξα
ˆ
f(ξ).
Exercise 24.9
Prove
F[e
iαx
f(x)] =
ˆ
f(x −α).
Exercise 24.10
Prove the following
F[cos (αx)f(x)] =
1
2
[
ˆ
f(ξ + α) +
ˆ
f(ξ −α)]
F[sin (αx)f(x)] =
1
2
[
ˆ
f(ξ + α) −
ˆ
f(ξ −α)]
216 THE FOURIER TRANSFORM SOLUTIONS FOR PDES
Sample Exam Questions
Exercise 24.11
Prove
F[f

(x)] = (iξ)
ˆ
f(ξ).
Exercise 24.12
Find the Fourier transform of f(x) = 1−|x| for −1 ≤ x ≤ 1 and 0 otherwise.
Exercise 24.13
Find, using the deﬁnition, the Fourier transform of
f(x) =
_
_
_
−1 −a < x < 0
1 0 < x < a
0 otherwise
Exercise 24.14
Find the inverse Fourier transform of
ˆ
f(ξ) = e

ξ
2
2
.
Exercise 24.15
Find F
−1
_
1
a+iξ
_
.
25 APPLICATIONS OF FOURIER TRANSFORMS TO PDES 217
25 Applications of Fourier Transforms to PDEs
Fourier transform is a useful tool for solving diﬀerential equations. In this
section, we apply Fourier transforms in solving various PDE problems. Con-
trary to Laplace transform, which usually uses the time variable, the Fourier
transform is applied to the spatial variable on the whole real line.
The Fourier transform will be applied to the spatial variable x while the vari-
able t remains ﬁxed. The PDE in the two variables x and t passes under the
Fourier transform to an ODE in the t−variable. We solve this ODE to obtain
the transformed solution ˆ u which can be converted to the original solution u
by means of the inverse Fourier transform. We illustrate these ideas in the
examples below.
First Order Transport Equation
Consider the initial value problem
u
t
+ cu
x
= 0
u(x, 0) = f(x).
Let ˆ u(ξ, t) be the Fourier transform of u in x. Performing the Fourier trans-
form on both the PDE and the initial condition, we reduce the PDE into an
ODE in t
∂ˆ u
∂t
+ iξcˆ u = 0
ˆ u(ξ, 0) =
ˆ
f(ξ).
Solution of the ODE gives
ˆ u(ξ, t) =
ˆ
f(ξ)e
−iξct
.
Thus,
u(x, t) = F
−1
[u(ξ, t)] = f(x −ct)
which is exactly the same as we obtained by using the method of character-
istics.
Second Order Wave Equation
Consider the two dimensional wave equation
u
tt
= c
2
u
xx
, x ∈ R, t > 0
218 THE FOURIER TRANSFORM SOLUTIONS FOR PDES
u(x, 0) = f(x)
u
t
(x, 0) = g(x).
Again, by performing the Fourier transform of u in x, we reduce the PDE
problem into an ODE problem in the variable t:

2
ˆ u
∂t
2
= −c
2
ξ
2
ˆ u
ˆ u(ξ, 0) =
ˆ
f(ξ)
ˆ u
t
(ξ, 0) = ˆ g(ξ).
General solution to the ODE is
ˆ u(ξ, t) = Φ(ξ)e
−iξct
+ Ψ(ξ)e
iξct
where Φ and Ψ are two arbitrary functions of ξ. Performing the inverse
transformation and making use of the translation theorem, we get the general
solution
u(x, t) = φ(x −ct) + ψ(x + ct)
where
ˆ
φ = Φ and
ˆ
ψ = Ψ. But
Φ(ξ) =
1
2
_
ˆ
f(ξ) −
1
iξc
ˆ g(ξ)
_
Ψ(ξ) =
1
2
_
ˆ
f(ξ) +
1
iξc
ˆ g(ξ)
_
.
By using the integration property, we ﬁnd the inverse transforms of Φ and Ψ
φ(x) =
1
2
_
f(x) +
1
c
_
x
0
g(s)ds
_
ψ(x) =
1
2
_
f(x) −
1
c
_
x
0
g(s)ds
_
.
Application of the translation property then yields directly the D’Alambert
solution
u(x, t) =
1
2
[f(x −ct) + f(x + ct)] +
1
2c
_
x+ct
x−ct
g(s)ds.
25 APPLICATIONS OF FOURIER TRANSFORMS TO PDES 219
Second Order Heat Equation
Next, we consider the heat equation
u
t
= ku
xx
, x ∈ R, t > 0
u(x, 0) = f(x).
Performing Fourier Transform in x for the PDE and the initial condition, we
obtain
∂ˆ u
∂t
= −kξ
2
ˆ u
ˆ u(ξ, 0) =
ˆ
f(ξ).
Treating ξ as a parameter, we obtain the solution to the above ODE problem
ˆ u(ξ, t) =
ˆ
f(ξ)e
−kξ
2
t
.
Application of the convolution theorem yields
u(x, t) =f(x) ∗ F
−1
]e
−kξ
2
t
]
=f(x) ∗
_
1

4πkt
e

x
2
4kt
_
=
1

4πkt
_

−∞
f(s)e

(x−s)
2
4kt
ds
Laplace’s Equation in 2D
Consider the problem
∆u = u
xx
+ u
yy
= 0, x ∈ R, 0 < y < L
u(x, 0) = 0
u(x, L) =
_
1 if −a < x < a
0 otherwise
Performing Fourier Transform in x for the PDE we obtain the second order
PDE in y
ˆ u
yy
= ξ
2
ˆ u.
The general solution is given by
ˆ u(ξ, y) = A(ξ) sinh ξy + B(ξ) cosh ξy.
220 THE FOURIER TRANSFORM SOLUTIONS FOR PDES
Using the boundary condition ˆ u(ξ, 0) = 0 we ﬁnd B(ξ) = 0. Using the second
boundary condition we ﬁnd
ˆ u(ξ, L) =
_

−∞
u(x, L)e
−iξx
dx
=
_
a
−a
e
−ξx
dx =
_
a
−a
cos ξxdx
=
2 sin ξa
ξ
Hence,
A(ξ) sinh ξL =
2 sin ξa
ξ
and this implies
A(ξ) =
2 sin ξa
ξ sinh ξL
.
Thus,
ˆ u(ξ, y) =
2 sin ξa
ξ sinh ξL
sinh ξy.
Taking inverse Fourier transform we ﬁnd
u(x, y) =
1

_

−∞
2 sin ξa
ξ sinh ξL
sinh ξye
iξx
dξ.
Noting that the integrand is an even function in ξ, we can simplify a little to
to obtain
u(x, y) =
1

_

−∞
2 sin ξa
ξ sinh ξL
sinh ξy cos ξxdξ
25 APPLICATIONS OF FOURIER TRANSFORMS TO PDES 221
Practice Problems
Exercise 25.1
Solve, by using Fourier transform
u
t
+ cu
x
= 0
u(x, 0) = e

x
2
4
.
Exercise 25.2
Solve, by using Fourier transform
u
t
= ku
xx
−αu, x ∈ R
u(x, 0) = e

x
2
γ
.
Exercise 25.3
Solve the heat equation
u
t
= ku
xx
subject to the initial condition
u(x, 0) =
_
1 if x ≥ 0
0 otherwise.
Exercise 25.4
Use Fourier transform to solve the heat equation
u
t
= u
xx
+ u, −∞< x < ∞< t > 0
u(x, 0) = f(x).
Exercise 25.5
Prove that
_

−∞
e
−|ξ|y
e
iξx
dξ =
2y
x
2
+ y
2
.
Exercise 25.6
Solve the Laplace’s equation in the half plane
u
xx
+ u
yy
= 0, −∞< x < ∞, 0 < y < ∞
subject to the boundary condition
u(x, 0) = f(x), |u(x, y)| < ∞.
222 THE FOURIER TRANSFORM SOLUTIONS FOR PDES
Exercise 25.7
Use Fourier transform to ﬁnd the transformed equation of
u
tt
+ (α + β)u
t
+ αβu = c
2
u
xx
where α, β > 0.
Exercise 25.8
Solve the initial value problem
u
t
+ 3u
x
= 0
u(x, 0) = e
−x
using the Fourier transform.
Exercise 25.9
Solve the initial value problem
u
t
= ku
xx
u(x, 0) = e
−x
using the Fourier transform.
25 APPLICATIONS OF FOURIER TRANSFORMS TO PDES 223
Sample Exam Questions
Exercise 25.10
Solve the initial value problem
u
t
= ku
xx
u(x, 0) = e
−x
2
using the Fourier transform.
Exercise 25.11
Solve the initial value problem
u
t
+ cu
x
= 0
u(x, 0) = x
2
using the Fourier transform.
Exercise 25.12
Solve, by using Fourier transform
∆u = 0
u
y
(x, 0) = f(x)
lim
x
2
+y
2
→∞
u(x, y) = 0.
224 THE FOURIER TRANSFORM SOLUTIONS FOR PDES
Section 1
1.1
(a) y
3
cos (xy)
(b) e
x
2
y
(2y + 4x
2
y
2
)
(c) 0
1.2
(a) f
x
(x, y) = 4x
3
, f
y
(x, y) =
3

y
(b) f
x
(x, y, z) = 2xy + 43, f
y
(x, y, z) = x
2
− 20yz
3

28
1+16y
2
, f
z
(x, y, z) =
−30y
2
z
2
(c) f
s
(s, t) =
2t
7
s

4
7
s

3
7
, f
t
(s, t) = 7t
6
ln (s
2
) −
27
t
4
(d) f
x
(x, y) =
4
x
2
sin
_
4
x
_
e
x
2
y−5y
3
+cos
_
4
x
_
e
x
2
y−5y
3
(2xy), f
y
(x, y) = cos
_
4
x
_
e
x
2
y−5y
3
(x
2

15y
2
)
(e) f
u
(u, v) =
9(u
2
+5v)−9u(2u)
(u
2
+5v)
2
=
−9u
2
+45v
(u
2
+5v)
2
, f
v
(u, v) =
−45u
(u
2
+5v)
2
(f) f
x
(x, y, z) =
sin y
z
2
, f
y
(x, y, z) =
xcos y
z
2
, f
z
(x, y, z) = −2
xsin y
z
3
(g) f
x
(x, y) =
1
2
_
2x +
5
5x−3y
2
_
(x
2
+ln (5x −3y
2
))

1
2
, f
y
(x, y) = −
3y
5x−3y
2
(x
2
+
ln (5x −3y
2
))

1
2
1.3 3
1.4
∂z
∂s
= t
2
e
st
2
sin (s
2
t) + 2ste
st
2
cos (s
2
t)
∂z
∂t
= 2ste
st
2
sin (s
2
t) + s
2
e
st
2
cos (s
2
t)
1.5 u is the depedent variable whereas x and y are the independent variables.
225
226 ANSWERS AND SOLUTIONS
1.6 We have
_
a
−a
f(x)dx =
_
0
−a
f(x)dx +
_
a
0
f(x)dx.
By the change of variable u = −x we ﬁnd
_
0
−a
f(u)du = −
_
0
a
f(−u)du = −
_
a
0
f(u)du.
Hence, the result follows.
1.7 We have
_
a
−a
f(x)dx =
_
0
−a
f(x)dx +
_
a
0
f(x)dx.
By the change of variable u = −x we ﬁnd
_
0
−a
f(u)du = −
_
0
a
f(−u)du =
_
a
0
f(u)du.
Hence, the result follows.
1.8 By the product rule of derivatives we have
(uv)

= u

v + uv

.
Integrate both sides to obtain
uv =
_
u

vdx +
_
uv

dx.
Now subtract
_
u

vdx from both sides to obtain the desired result.
1.9
u
tt
= −sin
_
x

_
sin
_
t

_
u
xx
= −sin
_
x

_
sin
_
t

_
.
1.10
u
tt
= sin
_
x

_
sinh
_
t

_
u
xx
= −sin
_
x

_
sinh
_
t

_
.
227
1.11
2
sup{
¸
¸
sinh
_
t

¸
}
1.12 (a) We have sup{|u
n
(x, 0) −1| : x ∈ R} =
1
n
sup{| sin nx| : x ∈ R} =
1
n
.
(b) We have sup{|u
n
(x, t) −1| : x ∈ R} =
e
n
2
t
n
Section 2
2.1 (a) For all 0 ≤ x < 1 we have lim
n→∞
f
n
(x) = lim
n→∞
x
n
= 0. Also,
lim
n→∞
f
n
(1) = 1. Hence, the sequence {f
n
}

n=1
converges pointwise to f.
(b) Suppose the contrary. Let =
1
2
. Then there exists a positive integer N
such that for all n ≥ N we have
|f
n
(x) −f(x)| <
1
2
for all x ∈ [0, 1]. In particular, we have
|f
N
(x) −f(x)| <
1
2
for all x ∈ [0, 1]. Choose (0.5)
1
N
< x < 1. Then |f
N
(x)−f(x)| = x
N
> 0.5 =
which is a contradiction. Hence, the given sequence does not converge uni-
formly.
2.2 For every real number x, we have
lim
n→∞
f
n
(x) = lim
n→∞
nx + x
2
n
2
= lim
n→∞
x
n
+ lim
n→∞
x
2
n
2
= 0
Thus, {f
n
}

n=1
converges pointwise to the zero function on R.
2.3 For every real number x, we have

1

n + 1
≤ f
n
(x) ≤
1

n + 1
.
Moreover,
lim
n]rightarrow∞
1

n + 1
= 0.
Applying the squeeze rule for sequences, we obtain
lim
n]rightarrow∞
f
n
(x) = 0
228 ANSWERS AND SOLUTIONS
for all x in R. Thus, {f
n
}

n=1
converges pointwise to the zero function on R.
2.4 First of all, observe that f
n
(0) = 0 for every n in N. So the sequence
{f
n
(0)}

n=1
is constant and converges to zero. Now suppose 0 < x < 1 then
n
2
x
n
= n
2
e
nln x
. But ln x < 0 when 0 < x < 1, it follows that
lim
n→∞
f
n
(x) = 0 for 0 < x < 1
Finally, f
n
(1) = n
2
for all n. So,
lim
n→∞
f
n
(1) = 1.
Therefore, {f
n
}

n=1
is not pointwise convergent on [0, 1].
2.5 For −
π
2
≤ x < 0 and 0 < x ≤
π
2
we have
lim
n→∞
(cos x)
n
= 0.
For x = 0 we have f
n
(0) = 1 for all n in N. Therefore, {f
n
}

n=1
converges
pointwise to
f(x) =
_
0 if −
π
2
≤ x < 0 and 0 < x ≤
π
2
1 if x = 0.
2.6 (a) Let > 0 be given. Let N be a positive integer such that N >
1

.
Then for n ≥ N
¸
¸
¸
¸
x −
x
n
n
−x
¸
¸
¸
¸
=
|x|
n
n
<
1
n

1
N
< .
Thus, the given sequence converges uniformly (and pointwise) to the function
f(x) = x.
(b) Since lim
n→∞
f

n
(x) = 1 for all x ∈ [0, 1), the sequence {f

n
}

n=1
converges
pointwise to f

(x) = 1. However, the convergence is not uniform. To see
this, let =
1
2
and suppose that the convergence is uniform. Then there is a
positive integer N such that for n ≥ N we have
|1 −x
n−1
−1| = |x|
n−1
<
1
2
.
In particular, if we let n = N + 1 we must have x
N
<
1
2
for all x ∈ [0, 1).
But x =
_
1
2
_ 1
N
∈ [0, 1) and x
N
=
1
2
N
<
1
2
. Hence, the
229
convergence is not uniform.
2.7 (a) The pointwise limit is
f(x) =
_
_
_
0 if 0 ≤ x < 1
1
2
if x = 1
1 if 1 < x ≤ 2
(b) The convergence cannot be uniform because if it were f would have to
be continuous.
2.8 (a) Let > 0 be given. Note that
|f
n
(x) −
1
2
| =
¸
¸
¸
¸
2 cos x −sin
2
x
2(2n + sin
2
x
¸
¸
¸
¸

3
4n
.
Since lim
n→∞
3
4n
=) we can ﬁnd a positive integer N such that if n ≥ N then
3
4n
< . Thus, for n ≥ N and all x ∈ R we have
|f
n
(x) −
1
2
| ≤
3
4n
< .
This shows that f
n

1
2
uniformly on R and also on [2, 7].
(b) We have
lim
n→∞
_
7
2
f
n
xdx =
_
7
2
lim
n→∞
f
n
xdx =
_
7
2
1
2
dx =
5
2
.
2.9 We have proved earlier that this sequence converges pointwise to the
discontinuous function
f(x) =
_
0 if −
π
2
≤ x < 0 and 0 < x ≤
π
2
1 if x = 0
Therefore, uniform convergence cannot occur for this given sequence.
2.10 (a) Using the squeeze rule we ﬁnd
lim
n→∞
sup{|f
n
(x)| : 2 ≤ x ≤ 5} = 0.
230 ANSWERS AND SOLUTIONS
Thus, {f
n
}

n=1
converges uniformly to the zero function.
(b) We have
lim
n→∞
_
5
2
f
n
(x)dx =
_
5
2
0dx = 0.
Section 3.
3.1 y =
1
2
(1 −e
−t
2
).
3.2 y(t) =
3t−1
9
+ e
−2t
+ Ce
−3t
.
3.3 y(t) = 3 sin t −
3 cos t
t
+
C
t
.
3.4 y(t) =
1
13
(3 sin (3t) + 2 cos (3t)) + Ce
−2t
.
3.5 y(t) = Ce
−sin t
−3.
3.6 α = −2.
3.7 p(t) = 2 and g(t) = 2t + 3.
3.8 y
0
= y(0) = −1 and g(t) = 2e
t
+ cos t + sin t.
3.9 1.
3.10 u(x, y) = f(bx −ay)e

c
a
2
+b
2
(ax+by)
.
3.11 y(t) = t ln |t| + 7t.
3.12 Since p(t) = a we ﬁnd µ(t) = e
at
. Suppose ﬁrst that a = λ. Then
y

+ ay = be
−at
and the corresponding general solution is
y(t) = bte
−at
+ Ce
−at
Thus,
lim
t→∞
y(t) = lim
t→∞
(
bt
e
at
+
C
e
at
)
= lim
t→∞
b
ae
at
= 0
231
Now, suppose that a = λ then
y(t) =
b
a −λ
e
−λt
+ Ce
−at
Thus,
lim
t→∞
y(t) = 0.
3.13 y(t) = (−te
t
+ e
t
)
−1
.
3.14 y(t) =
t
2
4

t
3
+
t
2
+
1
12t
2
.
3.15 y(t) = tSi(t) + (3 −Si(1))t.
Section 4
4.1 y(t) =
_
3
2
e
t
2
+ C
_1
3
.
4.2 y(t) = Ce
t
2
−2t
.
4.3 y(t) = Ct
2
+ 4.
4.4 y(t) =
2Ce
4t
1+Ce
4t
.
4.5 y(t) =
_
5 −4 cos (2t).
4.6 y(t) = −
_
(−2 cos t + 4).
4.7 y(t) = e
1−t
−1.
4.8 y(t) =
2

−8t+1
.
4.9 y(t) = tan (t + π) = tan t.
4.10 y(t) =
3−e
−2t
3+e
−2t
.
4.11 α =
1
2
, y
0
=
1
2
and n = 3.
4.12 u(x, y) = F(y)e
−3x
+ G(x) where F(y) =
_
f(y)dy.
232 ANSWERS AND SOLUTIONS
4.13 y
2
+ cos y + cos t +
t
2
2
= 2.
4.14 3y
2
y

+ cos y + 2t = 0, y(2) = 0.
4.15 The ODE is not separable.
Section 5
5.1 y(t) = e
t
−2e
3t
. lim
t→−∞
y(t) = 0 and lim
t→∞
y(t) = ∞.
5.2 y(t) = −2

2e
(−2−

2)t
+2

2e
(−2+

2)t
. lim
t→−∞
y(t) = −∞and lim
t→∞
y(t) =
0.
5.3 y(t) = −2e

2
2
t
. lim
t→−∞
y(t) = −∞ and lim
t→∞
y(t) = 0.
5.4 y

−y

−2y = 0.
5.5 y(t) = e
t
3
−1
(1 −t).
5.6 y(t) = e

2t
5
(t −1).
5.7 y(0) = 2 and y

(0) = −2.
5.8 y(t) = c
1
e
3t
+ c
2
te
3t
.
5.9 y(t) = 3e
−t
cos t + 2e
−t
sin t.
5.10 y(t) = −e
1
2
(t+π)
(3 cos
t
2
+ sin
t
2
).
5.11 y(t) = y
h
(t) + y
p
(t) = e
1
2
t
(c
1
cos

3
2
t + c
2
sin

3
2
t) +
6
73
cos 3t −
16
73
sin 3t.
5.12 y(t) = y
h
(t) + y
p
(t) = c
1
e
(−2−

6)t
+ c
2
e
(−2+

6)t
−t
2

5
2
t −9.
5.13 y = Ax
4
+ Bx
4
ln x.
5.14 y = x
−1
(Acos

3 ln x + Bsin

3 ln x).
233
5.15 (a) λ
n
= n
2
, y
n
(x) = sin nx, n = 1, 2, · · · .
(b) λ
n
=
__
n −
1
2
_
π
L
¸
2
and y
n
= sin
_
π
L
_
n −
1
2
_
x
_
, n = 1, 2, 3, · · · .
(c) λ
n
=
_
π
_
n −
1
2
__
2
, y
n
(x) = cos
_
π
_
n −
1
2
__
x, n = 1, 2, · · · .
5.16 We consider ﬁrst the cases (a) and (b). Multiply the equation by y

(x)
and integrate in x from 0 to L.
_
L
0
(ky

(x))

y(x)dx +
_
L
0
λy
2
(x)dx = 0.
Use integration by parts in the ﬁrst integral
[ky

(x)y(x)]
L
0

_
L
0
k(y

(x))
2
dx +
_
L
0
λy
2
(x)dx = 0.
The boundary term vanishes because of the boundary conditions. We solve
the above equation for λ and obtain
λ =
_
L
0
k(y

(x))
2
dx
_
L
0
y
2
(x)dx
≥ 0.
For the case (c), we repeat the above argument but by integrating from −L
to L.
5.17 y(t) = 2e
t
2
.
5.18 y(t) = c
1
+ c
2
e
t

1
10
cos (2t) +
1
5
sin (2t) + 5te
t
5.19 y(t) =
17
15
e
t
+
1
6
e
−2t

1
2
t −
1
4

3
20
sin 2t −
1
20
cos 2t.
Section 6
6.1 (a) ODE (b) PDE (c) ODE.
6.2 u
ss
= 0.
6.3 u
ss
+ u
tt
= 0.
6.4 (a) Order 3, nonlinear (b) Order 1, linear, homogeneous (c) Order 2,
234 ANSWERS AND SOLUTIONS
linear, inhomogeneous.
6.5 (a) Linear, homogeneous, order 3.
(b) Linear, inhomogeneous, order 4. The inhomogeneity is −sin y.
(c) Nonlinear, order 2. The nonlinear term is uu
x
.
(d) Nonlinear, order 3. The nonlinear terms are u
x
u
xxy
and uu
y
.
(e) Linear, inhomogeneous, order 2. The inhomogeneity is f(x, y, t).
6.6 (a) Linear. (b) Linear. (c) Nonlinear. (d) Nonlinear.
6.7 (a) PDE, linear, second order, homogeneous.
(b) PDE, linear, second order, homogeneous.
(c) PDE, nonlinear, fourth order.
(d) ODE, linear, second order, nonhomogeneous.
(e) PDE, linear, second order, nonhomogeneous.
(f) PDE, quasilinear, second order.
6.8 A(x, y, z)u
xx
+B(x, y, z)u
xy
+C(x, y, z)u
yy
+E(x, y, z)u
xz
+F(x, y, z)u
yz
+
G(x, y, z)u
zz
H(x, y, z)u
x
+I(x, y, z)u
y
+J(x, y, z)u
z
+K(x, y, z)u = L(x, y, z).
6.9 (a) Order 3, linear, homogeneous.
(b) Order 1, nonlinear.
(c) Order 4, linear, nonhomogeneous
(d) Order 2, nonlinear.
(e) Order 2, linear, homogeneous.
6.10 u
ww
= 0.
6.11 u
vw
= 0.
6.12 u
vw
= 0.
6.13 u
s
= 0.
6.14 u
s
=
1
2
.
6.15 u
w
= u.
235
Section 7
7.1 a = b = 0.
7.2 Subtituting into the diﬀerential equation we ﬁnd
tX

T −XT

= 0
or
X

X
=
T

tT
.
The LHS is a function of x only whereas the RHS is a function of t only.
This is true only when both sides are constant. That is, there is λ such that
X

X
=
T

tT
= λ
and this leads to the two ODEs X

= λX and T

= λtT.
7.3 We have xu
x
+ (x + 1)yu
y
=
x
y
(e
x
+ xe
x
) + (x + 1)y
_

xe
x
y
2
_
= 0 and
u(1, 1) = e.
7.4 We have u
x
+u
y
+2u = e
−2y
cos (x −y)−2e
−2y
sin (x −y)−e
−2y
cos (x −y)+
2e
−2y
sin (x −y) = 0 and u(x, 0) = sin x.
7.5 (a) The general solution to this equation is u(x) = C where C is an
arbitrary constant.
(b) The general solution is u(x, y) = f(y) where f is an arbitrary function of
y.
7.6 (a) The general solution to this equation is u(x) = C
1
x + C
2
where
C
1
and C
2
are arbitrary constants.
(b) We have u
y
= f(y) where f is an arbitrary function of y. Hence, u(x, y) =
_
y
a
f(t)dt.
236 ANSWERS AND SOLUTIONS
7.7 Let v(x, y) = y + 2x. Then
u
x
=2f
v
(v) + g(v) + 2xg
v
(v)
u
xx
=4f
vv
(v) + 4g
v
(v) + 4xg
v
(v)
u
y
=f
v
(v) + xg
v
(v)
u
yy
=f
vv
(v) + xg
vv
(v)
u
xy
=2f
vv
(v) + g
v
(v) + 2xg
vv
(v)
Hence,
u
xx
−4u
xy
+ 4u
yy
=4f
v
(v) + 4g
v
(v) + 4xg
v
(v)
−8f
vv
(v) −4g
v
(v) −8xg
vv
(v)
+4f
vv
(v) + 4xg
vv
(v) = 0.
7.8 u
tt
= c
2
u
xx
.
7.9 Let v = x + p(u)t. Using the chain rule we ﬁnd
u
t
= f
v
· v
t
= f
v
· (p(u) + p
u
u
t
t).
Thus
(1 −tf
v
p
u
)u
t
= f
v
p.
If 1 − tf
v
p
u
= 0 on any t−interval I then f
v
p = 0 on I which implies that
f
v
= 0 or p = 0 on I. But either condition will imply that tf
v
p
u
= 0 and
this will imply that 1 = 1 −tf
v
p
u
= 0, a contradiction. Hence, we must have
1 −tf
v
p
u
= 0. In this case,
u
t
=
f
v
p
1 −tf
v
p
u
.
Likewise,
u
x
= f
v
(1 + p
u
u
x
t)
or
u
x
=
f
v
1 −tf
v
p
u
.
It follows that u
t
= p(u)u
x
.
If u
t
= sin uu
x
then p(u) = sin u so that the general solution is given by
u(x, t) = f(x + sin ut)
237
where f is an arbitrary diﬀerentiable function in one variable.
7.10 u(x, y) = xf(x −y) + g(x −y).
7.11 Using integration by parts, we compute
_
L
0
u
xx
(x, t)u(x, t)dx = u
x
(x, t)u(x, t)|
L
x=0

_
L
0
u
2
(x, t)dx
=u
x
(L, t)u(L, t) −u
x
(0, t)u(0, t) −
_
L
0
u
2
(x, t)dx
=−
_
L
0
u
2
(x, t)dx ≤ 0
Note that we have used the boundary conditions u(0, t) = u(L, t) = 0 and
the fact that u
2
x
(x, t) ≥ 0 for all x ∈ [0, L].
7.12 (a) This can be done by plugging in the equations.
(b) Plug in.
(c) We have sup{|u
n
(x, 0) −1| : x ∈ R} =
1
n
sup{| sin nx| : x ∈ R} =
1
n
.
(d) We have sup{|u
n
(x, t) −1| : x ∈ R} =
e
n
2
t
n
.
(e) We have lim
t→∞
sup{|u
n
(x, t) − 1| : x ∈ R, t > 0} = lim
t→∞
e
n
2
t
n
= ∞.
Hence, the solution is unstable and thus the problem is ill-posed.
7.13 (a) u(x, y) = x
3
+ xy
2
+ f(y), where f is an arbitrary function.
(b) u(x, y) =
x
3
y
2
6
+ F(x) + g(y), where F(x) =
_
f(x)dx.
(c) u(x, t) =
1
18
e
2x+3t
.
7.14 (b) u(x, y) = xf(y −2x) + g(y −2x).
7.15 We have
u
t
=cu
v
−cu
w
u
tt
=c
2
u
vv
−2c
2
u
wv
+ c
2
u
ww
u
x
=u
v
+ u
w
u
xx
=u
vv
+ 2u
vw
+ u
ww
Substituting we ﬁnd u
vw
= 0 and solving this equation we ﬁnd u
v
= f(v)
and u(v, w) = F(v) + G(w) where F(v) =
_
f(v)dv.
238 ANSWERS AND SOLUTIONS
Finally, using the fact that v = x + ct and w = x − ct; we get d’Alembert’s
solution to the one-dimensional wave equation:
u(x, t) = F(x + ct) + G(x −ct)
where F and G are arbitrary diﬀerentiable functions.
Section 8
8.1 (a) Linear (b) Quasi-linear, nonlinear (c) Nonlinear (d) Semi-linear, non-
linear.
8.2 Let w = 2x −y. Then u
x
+ 2u
y
−u = e
x
f(w) + 2e
x
f
w
(w) −2e
x
f
w
(w) −
e
f
(w) = 0.
8.3 We have xu
x
− yu
y
= x(

xy +
xy
2

y
) − y
x
2

xy
= x

xy = u. Also,
u(y, y) = y
2
.
8.4 We have −yu
x
+xu
y
= −2xy sin (x
2
+ y
2
) +2xy sin (x
2
+ y
2
) = 0. More-
over, u(0, y) = cos y
2
.
8.5 We have
1
x
u
x
+
1
y
u
y
=
1
x
(−x)+
1
y
(1+y) =
1
y
. Moreover, u(x, 1) =
1
2
(3−x
2
).
8.6 3a −7b = 0.
8.7 Plug u = av + w into the equation. Using the linearity of L and the
assumptions on v and w, obtain
Lu = L(av + w) = aLv +Lw = 0 + f = f
for any constant a. Therefore, u solves the inhomogeneous equation for any a.
8.8 u
s
+
c
a
2
+b
2
= 0.
8.9 u(x, t) =
1
2
(x + y) + f(x −y).
239
8.10 Using integration by parts we ﬁnd
dI
dt
(t) =
_

−∞
∂u
∂t
dx =
_

−∞
xu
x
dx
= xf(x)|

−∞

_

−∞
udx
=−
_

−∞
udx ≤ 0
Hence, I(t) is decreasing.
8.11 We have
u
x
=−4e
−4x
f(2x −3y) + 2e
−4x
f

(2x −3y)
u
y
=−3e
−4x
f

(2x −3y)
Thus,
3u
x
+ 2u
y
+ 12u =−12e
−4x
f(2x −3y) + 6e
−4x
f

(2x −3y)
−6e
−4x
f

(2x −3y) + 12e
−4x
f(2x −3y) = 0.
8.12 u(x, y) = f(ax −bt)e
t
a
.
8.13 u(x, y) = f(bx −ay).
8.14 u
w
+ λu = f(v + cw, w).
8.15 vw
v
(v) = Aw(v).
Section 9
9.1 u(x, t) = sin (x −3t).
9.2 u(x, y) = e
c(ax+by)
a
2
+b
2
f(bx −ay).
9.3 u(x, y) = x cos (2x −y) + f(y −2x).
9.4 The change of coordinates v = x + t and w = x − t reduces the
240 ANSWERS AND SOLUTIONS
original equation to the equation u
v
=
v+w
2
whose solution is given by
u(v, w) =
v
2
4
+
wv
2
+g(w) or u(x, t) =
(x+t)
2
4
+
x
2
−t
2
2
+g(x−t). But u(x, x) = 1
so that 1 = x
2
+ g(0) or g(0) = x
2
− 1 which is impossible since g(0) is a
constant. Hence, the given initial value problem has no solution.
9.5 u(x, t) =
e
−3t
1+(x−t)
2
.
9.6 u(x, t) = e
3t
_
(x −t)
2
+
1
9
¸

1
3
t −
1
9
.
9.7 Using the chain rule we ﬁnd w
t
= u
t
e
λt
+ λue
λt
and w
x
= u
x
e
λt
. Substi-
tuting these equations into the original equation we ﬁnd
w
t
e
−λt
−λu + w
x
e
−λt
+ λu = 0
or
w
t
+ w
x
= 0
9.8 u(x, y) =
h(x−y)
1−yh(x−y)
.
9.9 (a) w(x, t) is a solution to the equation follows from the principle of
superposition. Moreover, w(x, 0) = u(x, 0) −v(x, 0) = f(x) −g(x).
(b) w(x, t) = f(x −ct) −g(x −ct).
(c) From (b) we see that
max
x,t
{|u(x, t) −v(x, t)|} = max
x
{|f(x) −g(x)|}.
Thus, small changes in the initial data produces small changes in the solu-
tion. Hence, the problem is a well-posed problem.
9.10
u(x, t) =
_
g
_
t −
x
c
_
e

λ
c
x
if x < ct
0 if x > ct.
9.11 u(x, t) = −sin
_
3t−2x
2
_
.
9.12 u(x, y) =
1
2
(x + y) + f(x −y).
9.13 (a) a = 1, b = 0, c = B, and d = −A (b) u(x, y) = f(Bx −Ay)e

C
A
x
.
241
9.14 u(x, y) = f(x −y)e
−x
.
9.15 u(x, y) = f(x −y)e
−x
+ x + y −2.
Section 10
10.1 The characteristics are hyperbolas: xy = k.
10.2 The characteristics are circles centered at the origin: x
2
+ y
2
= k.
10.3 The characteristics are parallel lines with common slope equals to
1 : x −y = k.
10.4 f
_
y
x
,
1
2
xy −arctan u
_
= 0 where f is an arbitrary diﬀerentiable function.
10.5 f(y + u, y ln (y + u) − x) or u = −yg(y ln (y + u) − x) where f and
g are arbitrary diﬀerentiable functions.
10.6 f
_
y
x
,
u
x
_
= 0 or u = xg
_
y
x
_
where f and g are arbitrary diﬀerentiable
functions.
10.7 f
_
y
x
,
u
x
n
_
= 0 or u = x
n
g
_
y
x
_
where f and g are arbitrary diﬀeren-
tiable functions.
10.8 f(x + y + z,
x
uy
) = 0 where f is an arbitrary diﬀerentiable function.
10.9 f(xy, x
4
−u
4
−2xyu
2
) = 0, where f is an arbitrary diﬀerentiable func-
tion.
10.10 f(x
2
+ y
2
− u
2
, xy + u) = 0 where f is an arbitrary diﬀerentiable
function.
10.11 f(
y
u
, x
2
+ y
2
+ u
2
) = 0 where f is an arbitrary diﬀerentiable func-
tion.
10.12 u(x, y) = e
x
f(y −2x).
10.13 u(x, y) = f(y − arctan x) for any diﬀerentiable function f. The char-
242 ANSWERS AND SOLUTIONS
acteristics are shown below.
10.14 u = e
x
f(ye
−x
) where f is an arbitrary diﬀerential function.
10.15 The characteristics are solutions to the DE
dy
dx
= x. Solving this ODE
we ﬁnd y =
x
2
2
+ C.
10.16 u = e

x
2
2
f(ye
−x
) where f is an arbitrary diﬀerentiable function of
one variable.
10.17 Solving
dy
dx
=
x
y
by the separation of variables we ﬁnd x
2
− y
2
= k,
where k is a constant.
10.18 Solving
dy
dx
=
x
y
by the separation of variables we ﬁnd x
2
− y
2
= k,
where k is a constant.
Section 11
11.1 u(x, y) =
1−xy
x+y
, x + y = 0.
11.2 u(x, y) = (x + y)(x
2
−y
2
).
11.3 2xyu + x
2
+ y
2
−2u + 2 = 0.
11.4 u(x, y) = ln
_
x + 1 −
y
x
_
.
243
11.5 u(x, y) = f(xe
−y
).
11.6 u(t, x) = f(x −at).
11.7 u(x, y) =
1
sec (x−ay)−y
.
11.8 u(x, y) = h
_
y −
(x−1)
2
2
−(x −1)
_
e
x−1
.
11.9 u(x, y) = f(x −uy).
11.10 u(x, y) = y −sin
−1
x.
11.11 (i) y = Cx
2
. The characteristics are parobolas in the plane centered
at the origin. See ﬁgure below.
(ii) u(x, y) = e
yx
−2
.
(iii) In the ﬁrst case, we cannot substitute x = 0 into yx
−2
(the argument
of the function f, above) because x
−2
is not deﬁned at 0. Similarly, in the
second case, we’d need to ﬁnd a function f so that f(0) = h(x). If h is not
constant, it is not possible to satisfy this condition for all x ∈ R.
(iv) All characteristics intersect at (0, 0). Since the solution is constant along
any characteristic, if the solution is not exactly constant for all (x, y), then
the limit of u(x, y) as (x, y) → (0, 0) is diﬀerent if we approach (0, 0) along
diﬀerent characteristics. Therefore, the method doesn’t work at that point.
244 ANSWERS AND SOLUTIONS
11.12 u(x, y) = e
y
cos (x −y).
11.13 (a) u = e
x
f(ye
−x
) where f is an arbitrary diﬀerential function.
(b) We want 2 = u(x, 3x) = e
x
f(3e
x
e
−x
) = e
x
f(3). This equation is impossi-
ble so this Cauchy problem has no solutions.
(c) We want e
x
= e
x
f(e
x
e
−x
) =⇒ f(1) = 1. In this case, there are inﬁnitely
many solutions to this Cauchy problem, namely, u(x, y) = e
x
f(ye
−x
) where
f is an arbitrary function satisfying f(1) = 1.
11.14 u(x, y) = −1 + 2e
x
2
2
e

(4x−y)
2
2
.
11.15 The Cauchy problem has no solutions.
11.16 (a) The characteristics satisfy the ODE
dy
dx
=
x
y
. Solving this equa-
tion we ﬁnd x
2
−y
2
= C. Thus, the characteristics are hyperbolas.
(b)
(c) The general solution to the PDE is u(x, y) = f(x
2
− y
2
) where f is
an arbitrary diﬀerentiable function. Since u(0, y) = e
−y
2
we ﬁnd f(y) = e
y
.
Hence, u(x, y) = e
x
2
−y
2
.
(d) This solution is only deﬁned in the region covered by characteristics that
cross the y axis: y
2
−x
2
> 0. The solution in the region y
2
−x
2
< 0 can be
245
any function of the form u(x, y) = f(x
2
−y
2
).
11.17 (a) Solving the ODE
dy
dx
= y we ﬁnd the characteristics ye
−x
= C.
Thus, u(x, y) = f(ye
−x
). If u(x, 0) = 1 then we choose f to be any arbitrary
diﬀerentiable function satisfying f(0) = 1.
(b) The line y = 0 is a characteristic so that u has to be constant there.
Hence, there is no solution satisfying the condition u(x, 0) = x.
Section 12
12.1 (a) Hyperbolic (b) Parabolic (c) Elliptic.
12.2 (a) Ellitpic (b) Hyperbolic (c)Hyperbolic.
12.3 • The PDE is of hyperbolic type if 4y
2
(x
2
+x +1) > 0. This is true for
all y = 0 and x
2
+ x + 1 > 0. Graphically, this is the inside of the parabola
x
2
+ x + 1 that opens up.
• The PDE is of parabolic type if 4y
2
(x
2
+ x + 1) = 0. Since x
2
+ x + 1 > 0
for all x ∈ R, we must have y = 0. Graphically, this is x−axis.
• The PDE is of elliptic type if 4y
2
(x
2
+ x + 1) < 0 which can not happen.
12.4 We have
u
x
(x, t) = −sin x sin t,
u
xx
(x, t) = −cos x sin t,
u
t
(x, t) = cos x cos t,
u
tt
(x, t) = −cos x sin t.
Thus,
u
xx
(x, t) = −cos x sin t = u
tt
(x, t),
u(x, 0) = −sin x sin 0 = 0,
u
t
(x, 0) = cos x cos 0 = cos x,
u
x
(0, t) = −sin 0 sin t = 0.
12.5 (a) Quasi-linear (b) Semi-linear (c) Linear (d) Nonlinear.
246 ANSWERS AND SOLUTIONS
12.6 We have
u
x
=
2x
x
2
+ y
2
u
xx
=
2y
2
−2x
2
(x
2
+ y
2
)
2
u
y
=
2y
x
2
+ y
2
u
yy
=
2x
2
−2y
2
(x
2
+ y
2
)
2
Plugging these expressions into the equation we ﬁnd u
xx
+ u
yy
= 0. Similar
argument holds for the second part of the problem.
12.7 Multiplying the equation by u and integrating, we obtain
λ
_
L
0
u
2
(x)dx =
_
L
0
uu
xx
(x)dx
=[u(L)u
x
(L) −u(0)u
x
(0)] −
_
L
0
u
2
x
(x)dx
=−
_
k
L
u
2
x
(L) + k
0
u
(
x
0) +
_
L
0
u
2
x
(x)dx
_
For λ > 0, because k
0
, k
L
> 0, the right-hand side is nonpositive and the
left-hand side is nonnegative. Therefore, both sides must be zero, and there
can be no solution other than u ≡ 0, which is the trivial solution.
12.8 Substitute u(x, y) = f(x)g(y) into the left side of the equation to obtain
f(x)g(y)(f(x)g(y))
xy
= f(x)g(y)(f

(x)g

(y)). Now, substitute the same thing
into the right side to obtain (f(x)g(y))
x
(f(x)g(y))
y
= f

(x)g(y)f(x)g

(y) =
f(x)g(y)f

(x)g

(y). So the sides are equal, which means f(x)g(y) is a solution.
12.9 We have
u
xx
= −n
2
sin nx sinh ny and u
yy
= n
2
sin nx sinh ny
Hence, ∆u = 0.
12.10 u(x, y) =
x
2
y
2
4
+ F(x) + G(y), where F(x) =
_
f(x)dx.
247
12.11 (a) We have A = 2, B = −4, C = 7 so B
2
−4AC = 16−56 = −40 < 0.
So this equation is elliptic everywhere in R
2
.
(b) We have A = 1, B = −2 cos x, C = −sin
2
x so B
2
− 4AC = 4 cos
2
x +
4 sin
2
x = 4 > 0. So this equation is hyperbolic everywhere in R
2
.
(c) We have A = y, B = 2(x − 1), C = −(y + 2) so B
2
− 4AC =
4(x −1)
2
+ 4y(y + 2) = 4[(x −1)
+
(y + 1)
2
−4]. The equation is parabolic if
(x −1)
2
+(y +1)
2
= 4. It is hyperbolic if (x −1)
2
+(y +1)
2
> 4 and elliptic
if (x −1)
2
+ (y + 1)
2
< 4.
12.12 Using the chain rule we ﬁnd
u
t
(x, t) =
1
2
(cf

(x + ct) −cf

(x −ct)) +
1
2c
[g(x + ct)(c) −g(x −ct)(−c))
=
c
2
(f

(x + ct) −f

(x −ct)) +
1
2
(g(x + ct) + g(x −ct))
u
tt
=
c
2
2
(f

(x + ct) + f

(x −ct)) +
c
2
(g

(x + ct) −g

(c −xt))
u
x
(x, t) =
1
2
(f

(x + ct) + f

(x −ct)) +
1
2c
[g(x + ct) −g(x −ct))
u
xx
(x, t) =
1
2
(f

(x + ct) + f

(x −ct)) +
1
2c
[g

(x + ct) −g

(x −ct))
By substitutition we see that c
2
u
xx
= u
tt
. Moreover,
u(x, 0) =
1
2
(f(x) + f(x)) +
1
2c
_
x
x
g(s)ds = f(x)
and
u
t
(x, 0) = g(x).
12.13 (a) 1 + 4x
2
y > 0, (b) 1 + 4x
2
y = 0, (c) 1 + 4x
2
y = 0.
12.14 u(x, y) = f(y −3x) + g(x + y).
12.15 u(x, y) = f(y −3x) + g(x + y) =
10x
2
+y
2
−7xy+6
6
.
Section 13
248 ANSWERS AND SOLUTIONS
13.1 Let z(x, t) = αv(x, t) + βw(x, t). Then we have
c
2
z
xx
=c
2
αv
xx
+ c
2
βw
xx
=αv
tt
+ βv
tt
=z
tt
.
13.2 Indeed we have c
2
u
xx
(x, t) = 0 = u
tt
(x, t).
13.3 u(x, t) = 0.
13.4 u(x, t) =
1
2
(cos (x −3t) + cos (x + 3t)).
13.5 u(x, t) =
1
2
_
1
1+(x+t)
2
+
1
1+(x−t)
2
_
.
13.6 u(x, t) = 1 +
1

[sin (2π + 4πt) −sin (2π −4πt)].
13.7
u(x, t) =
_
_
_
1 if x −5t < 0 and x + 5t < 0
1
2
if x −5t < 0 and x + 5t > 0
0 if x −5t > 0.
13.8 u(x, t) =
1
2
[e
−(x+ct)
2
+ e
−(x−ct)
2
] +
t
2
+
1
4c
cos (2x) sin (2ct).
13.9 Just plug the translated/diﬀerentiated/dialated solution into the wave
equation and check that it is a solution.
13.10 v(r) = Acos nr + Bsin nr.
13.11 u(x, t) =
1
2
[e
x−ct
+ e
x+ct
+
1
c
(cos (x −ct) −cos (x + ct))].
13.12 (a) We have
dE
dt
(t) =
_
L
0
(u
t
u
tt
+
_
L
0
c
2
u
x
u
xt
)dx
=
_
L
0
u
t
u
tt
dx + c
2
u
t
(L, t)u
x
(L, t) −c
2
u
t
(0, t)u
x
(0, t) −c
2
_
L
0
u
t
u
xx
dx
=c
2
u
t
(L, t)u
x
(L, t) −c
2
u
t
(0, t)u
x
(0, t) +
_
L
0
u
t
(u
tt
−c
2
u
xx
)dx
=c
2
(u
t
(L, t)u
x
(L, t) −u
t
(0, t)u
x
(0, t))
249
since u
tt
−c
2
u
xx
= 0.
(b) Since the ends are ﬁxed, we have u
t
(0, t) = u
t
(L, t) = 0. From (a) we
have
dE
dt
(t) = c
2
(u
t
(L, t)u
x
(L, t) −u
t
(0, t)u
x
(0, t)) = 0.
(c) Assuming free ends boundary conditions, that is u
x
(0, t) = u
x
(L, t) = 0,
we ﬁnd
dE
dt
(t) = 0.
13.13 Using the previous exercise, we ﬁnd
dE
dt
(t) = −d
_
L
0
(u
t
)
2
dx.
The right-hand side is nonpositive, so the energy either decreases or is con-
stant. The latter case can occur only if u
t
(x, t) is identically zero, which
means that the string is at rest.
13.14 (a) By the chain rule we have u
t
(x, t) = −cR

(x − ct) and u
tt
(x, t) =
c
2
R

(x − ct). Likewise, u
x
(x, t) = R

(x − ct) and u
xx
= R

(x − ct). Thus,
u
tt
= c
2
u
xx
.— (b) We have
_
L
0
(u
t
)
2
dx =
_
L
0
c
2
2
[R

(x −ct)]
2
dx =
_
L
0
c
2
2
(u
x
)
2
dx.
13.15 u(x, t) = x
2
+ 4t
2

1
4
sin 2x sin 4t.
Section 14
14.1 Let z(x, t) = αu(x, t) + βv(x, t). Then we have
kz
xx
=kαu
xx
+ kβv
xx
=αu
t
+ βv
t
=z
t
.
14.2 Indeed we have ku
xx
(x, t) = 0 = u
t
(x, t).
14.3 u(x, t) = T
0
+
T
L
−T
0
L
x.
14.4 Let u be the solution to (14.1) that satisﬁes u(0, t) = u(L, t) = 0. Let
250 ANSWERS AND SOLUTIONS
w(x, t) be the time independent solution to (14.1) that satisﬁes w(0, t) = T
0
and w(L, t) = T
L
. That is, w(x, t) = T
0
+
T
L
−T
0
L
x. From Exercise 14.1,
the function u(x, t) = u(x, t) + w(x, t) is a solution to (14.1) that satis-
ﬁes u(0, t) = T
0
and u(L, t) = T
L
.
14.5 u(x, t) = 0.
14.6 Substituting u(x, t) = X(x)T(t) into (14.1) we obtain
k
X

X
=
T

T
.
Since X only depends on x and T only depends on t, we must have that
there is a constant λ such that
k
X

X
= λ and
T

T
= λ.
This gives the two ordinary diﬀerential equations
X

λ
k
X = 0 and T

−λT = 0.
14.7 (a) Letting α =
λ
k
> 0 we obtain the ODE X

−αX = 0 whose general
solution is given by X(x) = Ae

αx
+ Be

αx
for some constants A and B.
(b) The condition u(0, t) = 0 implies that X(0) = 0 which in turn implies
A+B = 0. Likewise, the condition u(L, t) = 0 implies Ae

λL
+Be

λL
= 0.
Hence, A(e

λL
−e

λL
) = 0.
(c) If A = 0 then B = 0 and u(x, t) is the trivial solution which contradicts
the assumption that u is non-trivial. Hence, we must have A = 0.
(d) Using (b) and (c) we obtain e

λL
= e

λL
. Since L > 0, the number
on the left is greater than 1 whereas the number on the right is less than
1. Hence, we must have A = B = 0 which leads to the trivial solution, a
contradiction. Hence, we must have λ < 0.
14.8 (a)Now, write σ = −λ > 0. Then we obtain the equation X

+
σ
k
X = 0
whose general solution is given by
X(x) = c
1
cos βx + c
2
sin βx
where β =
_
σ
k
=
_
−λ
k
.
(b) Using X(0) = 0 we obtain c
1
= 0. Since c
2
= 0 we must have sin βL = 0.
251
Thus, σ =
kn
2
π
2
L
2
, where n is an integer.
14.9 For each integer n ≥ 0 we have u
n
(x, t) =
cn
T(0)
T(0)e
kn
2
π
2
L
2
t
sin
_

L
_
x
is a solution to (14.1). By superposition, u(x, t) is also a solution to (14.1).
Moreover, u(0, t) = u(L, t) = 0 since u
n
(0, t) = u
n
(L, t) = 0.
14.10 (i) u(0, t) = 0 and u(a, t) = 100 for t > 0.
(ii) u
t
(0, t) = u
t
(a, t) = 0 for t > 0.
14.11 Solving this problem we ﬁnd u(x, t) = e
−t
sin x. We have
E(t) =
_
π
0
[e
−2t
sin
2
x + e
−2t
cos
2
x]dx =
_
π
0
e
−2t
dt = πe
−2t
.
Thus, E

(t) = −2πe
2
t
< 0 for all t > 0.
14.12 E(t) =
_
L
0
f(x)dx + (1 + 4L)t.
14.13 v(x) = x + 2.
14.14 (a) v(x) =
T
L
x.
(b) v(x) = T.
(c) v(x) = αx + T.
(d) v(x) = −
x
2
2
+
_
T
2
−T
1
L
+
L
2
_
x + T
1
.
14.15 (a) We ﬁrst consider the intervals 0 < x < L and L < x < 2L
independently, and denote by u
1
(x) and u
2
(x) the temperature distributions
in the respective intervals. Thus,
u(x) =
_
u
1
(x) 0 < x < L
u
2
(x) L < x < 2
The steady temperature satisﬁes the one-dimensional Laplace equation, so
we have
u

1
= 0, 0 < x < L; u

2
= 0, L < x < 2L.
Solving the equations, we ﬁnd
u
1
(x) = C
1
x + C
2
, 0 < x < L; u
2
(x) = C
3
x + C
4
, L < x < 2L.
252 ANSWERS AND SOLUTIONS
We have the following conditions to determine the constants:
u
1
(0) = T
1
, u
2
(2L) = T
2
(boundary conditions)
u
1
(L) = u
2
(L) (continuity of the temperature)
k
1
u

1
(L) = k
2
u

2
(L) (continuity of the heat ﬂux)
From the ﬁrst two conditions,
u
1
(x) = T
1
+ C
1
x, u
2
(x) = T
2
+ C
3
(x −2L).
From the last condition,
k
1
C
1
= k
2
C
3
=⇒C
3
=
k
1
k
2
C
1
.
Finally, from the continuity of the temperature,
T
1
+ C
1
L = T
2
−C
3
L =⇒C
1
+ C
3
=
T
2
−T
1
L
,
and we obtain
C
1
=
k
2
k
1
+ k
2
·
T
2
−T
1
L
, C
3
=
k
1
k
1
+ k
2
·
T
2
−T
1
L
.
Thus,
u
1
(x) = T
1
+ (T
2
−T
1
)
k
2
k
1
+ k
2
x
L
, u
2
(x) = T
2
+ (T
2
−T
1
)
k
1
k
1
+ k
2
x −2L
L
.
(b) If L = 1, T
1
= 0, T
2
= 100, k
1
= 2, k
2
= 1 we ﬁnd
u
1
(x) =
100
3
x, u
2
(x) =
200
3
x −
100
3
.
The temperature is a piecewise linear function which has slope
100
3
for 0 <
x < 1 and slope
200
3
for 1 < x < 2.
14.16 (a) E(t) =
_
L
0
cρu(x, t)dx.
(b) We integrate the equation in x from 0 to L :
_
L
0
cρu
t
(x, t)dx =
_
L
0
Ku
xx
dx = Ku
x
(x, t)|
L
0
= 0,
253
since u
x
(0, t) = u
x
(L, t) = 0. The left-hand side can also be written as
d
dt
_
L
0
cρu(x, t)dx = E

(t).
Thus, we have shown that E

(t) = 0 so that E(t) is constant.
14.17 (a) The total thermal energy is
E(t) =
_
L
0
u(x, t)dx.
We have
dE
dt
=
_
L
0
u
t
(x, t)dx = u
x
|
L
0
+
_
L
0
xdx = (7 −β) +
L
2
2
.
(b) The steady solution (equilibrium) is possible if the right-hand side van-
ishes:
(7 −β) +
L
2
2
= 0
Solving this equation for β we ﬁnd β = 7 +
L
2
2
.
(c) By integrating the equation u
xx
+ x = 0 we ﬁnd the steady solution
u(x) = −
x
3
6
+ C
1
x + C
2
From the condition u
x
(0) = β we ﬁnd C
1
= β. The steady solution should
also have the same value of the total energy as the initial condition. This
means
_
L
0
_

x
3
6
+ βx + C
2
_
dx =
_
L
0
f(x)dx = E(0).
Performing the integration and then solving for C
2
we ﬁnd
C
2
=
1
L
_
L
0
f(x)dx +
L
3
24
−β
L
2
.
Therefore, the steady-state solution is
u(x) =
1
L
_
L
0
f(x)dx +
L
3
24
−β
L
2
+ βx −
x
3
6
.
254 ANSWERS AND SOLUTIONS
Section 15
15.1 (a) We have (fg)(x + T) = f(x + T)g(x + T) = f(x)g(x) = (fg)(x).
(b) We have (c
1
f +c
2
g)(x+T) = c
1
f(x+T)+c
2
g(x+T) = c
1
f(x)+c
2
g(x) =
(c
1
f + c
2
g)(x).
15.2 (a) For n = m we have
_
L
−L
sin
_

L
x
_
sin
_

L
x
_
dx =−
1
2
_
L
−L
_
cos
_
(m + n)π
L
x
_
−cos
_
(m−n)π
L
x
__
dx
=−
1
2
_
L
(m + n)π
sin
_
(m + n)π
L
x
_

L
(m−n)π
sin
_
(m−n)π
L
x
__
L
−L
=0
where we used the trigonometric identiy
sin a sin b =
1
2
[−cos (a + b) + cos (a −b)].
(b) For n = m we have
_
L
−L
cos
_

L
x
_
sin
_

L
x
_
dx =
1
2
_
L
−L
_
sin
_
(m + n)π
L
x
_
−sin
_
(m−n)π
L
x
__
dx
=−
1
2
_
L
(m + n)π
cos
_
(m + n)π
L
x
_
+
L
(m−n)π
cos
_
(m−n)π
L
x
__
L
−L
=0
where we used the trigonometric identiy
cos a sin b =
1
2
[sin (a + b) −sin (a −b)].
15.3 (a) L (b) L (c) 0.
255
15.4
a
0
=
1
π
_
π
−π
f(x)dx = 0
a
n
=
1
π
_
π
−π
f(x) cos nxdx
=−
_
0
−π
cos nxdx +
_
π
0
cos nxdx = 0
b
n
=
1
π
_
π
−π
f(x) sin nxdx
=−
_
0
−π
sin nxdx +
_
π
0
sin nxdx
=
2
n
[1 −(−1)
n
]
15.5 f(x) = −
1
6
+

n=1
4
(nπ)
2
(−1)
n
cos (nx).
15.6 f(x) =

n=1
2

_
cos
_

2
_
−(−1)
n
¸
sin
_
nx
2
_
.
15.7 f(x) =

n=1
4
(nπ)
2
[1 −(−1)
n
] cos
_

2
x
_
.
15.8 Since the sided limits at the point of discontinuity x = 0 do not exist,
the function is not piecewise continuous in [−1, 1].
15.9 Deﬁne the function
g(a) =
_
L+a
−L+a
f(x)dx.
Using the fundamental theorem of calculus, we have
dg
da
=
d
da
_
L+a
−L+a
f(x)dx
=f(L + a) −f(−L + a) = f(−L + a + 2L) −f(−L + a)
=f(−L + a) −f(−L + a) = 0
Hence, g is a constant function, and in particular we can write g(a) = g(0)
for all a ∈ R which gives the desired result.
256 ANSWERS AND SOLUTIONS
15.10 (i) f(x) =
10
3
+

n=1
[−
1

sin
_
2nπ
3
_
cos
_
2nπx
3
_
+(−
1

_
−cos
_
2nπ
3
_
+ 1
_
) sin
_
2nπx
3
_
.
(ii) Using the theorem discussed in class, because this function and its deriva-
tive are piecewise continuous, the Fourier series will converge to the function
at each point of continuity. At any point of discontinuity, the Fourier series
will converge to the average of the left and right limits.
(iii)
.
15.11 (a) a
0
= 2, a
n
= b
n
= 0 for n ∈ N.
(b) a
0
= 4, a
n
= 0, b
1
= 1, and b
n
= 0.
(c) a
0
= 1, a
n
= 0, b
n
=
1
πn
[1 −(−1)
n
], n ∈ N.
(d) a
0
= a
n
= 0, b
n
=
2L
πn
(−1)
n+1
, n ∈ N.
15.12 −1
15.13 a
n
= 0 for all n ∈ N.
15.14
f(0

)+f(0
+
)
2
=
−π+π
2
= 0.
15.15 (a) f(x) =
3
2
+
2
π

n=1
sin (2n−1)x
2n−1
.
(b)

n=1
(−1)
n+1
2n−1
=
π
4
.
Section 16
16.1 f(x) = 0.
16.2
257
16.3
16.4
258 ANSWERS AND SOLUTIONS
16.5 f(x) =
π
4
+

n=1
2
πn
2
[2 cos (nπ/2) −1 −(−1)
n
] cos nx.
16.6 f(x) =
π
2
+

n=1
2
n
2
π
[(−1)
n
−1] cos nx.
16.7 f(x) =

n=1
2

[1 −(−1)
n
] sin nx.
16.8 f(x) =
2
pi

n=1
n
_
1+(−1)
n
n
2
−1
_
sin nx.
16.9 f(x) =
1
2
(e
2
−1) +

n=1
4
4+n
2
π
2
cos (nπx).
16.10 (i) The formulas for the cosine and sine series of a function f(x)
on an interval [0, 2] are, respectively
a
0
2
+

n=1
a
n
cos
_
nπx
2
_
, a
n
=
_
2
0
x cos
_
nπx
2
_
dx

n=1
b
n
sin
_
nπx
2
_
, b
n
=
_
2
0
x sin
_
nπx
2
_
dx.
(ii) By the theorem discussed in class, since the function is continuously dif-
ferentiable , the Fourier sine series will converge pointwise to f on the open
259
interval (0, 2).
(iii) By the theorem discussed in class, since the function is continuously
diﬀerentiable , the Fourier cosine series will converge pointwise to f on the
open interval (0, 2).
(iv) The key point is that the even extension is continuous on [−2; 2], and so
its cosine series will be pointwise convergent everywhere. However, its odd
extension has a jump discontinuity at x = ±2, so there the sine series will
converge to the average of its left and right limit, which is zero.
The graph of the even extension is shown on the left, and that of odd exten-
sion is shown on the right.
On the left, the limit of the cosine series, which converges pointwise to the
even extension on the entire closed interval. On the right, the limit of the sine
series, which converges pointwise to the odd extension everywhere except at
x = ±2, where it converges to 0, the average of the left and right limits of
the odd extension there.
260 ANSWERS AND SOLUTIONS
16.11 (a) If f(x) = sin
_

L
x
_
then b
n
= 0 of n = 2 and b
2
= 1.
(b) If f(x) = 1 then
b
n
=
2
L
_
L
0
sin
_

L
x
_
dx =
2

[1 −(−1)
n
].
(c) If f(x) = cos
_
π
L
x
_
then
b
1
=
2
L
_
L
0
cos
_
π
L
x
_
sin
_
π
L
x
_
dx = 0
and for n = 1 we have
b
n
=
2
L
_
L
0
cos
_
π
L
x
_
sin
_
π
L
x
_
dx
=
1
2
2
L
_
L
0
_
sin
_
πx
L
_
(1 + n) −sin
_
πx
L
_
(1 −n)
_
dx
=
1
L
_

L
(1 + n)π
cos
_
πx
L
_
(1 + n) +
L
1 −n
cos
_
πx
L
_
(1 −n)
_
L
0
=
2n
(n
2
−1)π
[1 + (−1)
n
].
16.12 (a) a
0
= 10 and a
1
= 1, and a
n
= 0 for n = 1.
(b) a
0
= L and a
n
=
2L
(πn)
2
[(−1)
n
−1], n ∈ N.
(c) a
0
= 1 and a
n
=
2
πn
sin
_
πn
2
_
, n ∈ N.
16.13 By deﬁnition of Fourier sine coeﬃcients,
b
n
=
2
L
_
L
0
f(x) sin
_

L
x
_
dx
The symmetry around x =
L
2
can be written as
f
_
L
2
+ y
_
= f
_
L
2
−y
_
for all y ∈ R. To use this symmetry it is convenient to make the change of
variable x =
L
2
+ y in the above integral to obtain
b
n
=
_ L
2

L
2
f
_
L
2
+ y
_
sin
_

L
_
L
2
+ y
__
dx.
261
Since f
_
L
2
+ y
_
is even in y and for n even sin
_

L
_
L
2
+ y

= sin
_
nπy
L
_
is
odd in y, the integrand of the above integral is odd in y for n even. Since
the intergral is from −
L
2
to
L
2
we must have b
n
= 0 for all n ∈ N.
16.14 By deﬁnition of Fourier cosine coeﬃcients,
a
n
=
2
L
_
L
0
f(x) cos
_

L
x
_
dx
The anti-symmetry around x =
L
2
can be written as
f
_
L
2
−y
_
= −f
_
L
2
+ y
_
for all y ∈ R. To use this symmetry it is convenient to make the change of
variable x =
L
2
+ y in the above integral to obtain
a
n
=
_ L
2

L
2
f
_
L
2
+ y
_
cos
_

L
_
L
2
+ y
__
dx.
Since f
_
L
2
+ y
_
is odd in y and for n even cos
_

L
_
L
2
+ y

= cos
_
nπy
L
_
is
even in y, the integrand of the above integral is odd in y for n even. Since
the intergral is from −
L
2
to
L
2
we must have a
n
= 0 for all n = 0, 1, 2, · · · .
16.15 sin
_
nπx
L
_
=
2
π

2
π

n=2
1+(−1)
n
n
2
−1
cos
_
nπx
L
_
.
16.16 (a)
262 ANSWERS AND SOLUTIONS
(b) a
0
=
2
2
_
2
0
f(x)dx = 3.
(c) We have
a
n
=
2
2
_
2
0
f(x) cos
_
nπx
2
_
dx
=
_
1
0
cos
_
nπx
2
_
dx +
_
2
1
2 cos
_
nπx
2
_
dx
=
2

sin
_
nπx
2
_
¸
¸
¸
¸
1
0
+ 2
2

sin
_
nπx
2
_
¸
¸
¸
¸
2
1
=−
2

sin
_

2
_
.
(d) b
n
= 0 since this is an even function.
(e)
f(x) =
3
2
+

n=1
_

2

sin
_

2
_
_
cos
_
nπx
2
_
.
Section 17
17.1 We look for a solution of the form u(x, y) = X(x)Y (y). Substituting in
the given equation, we obtain
X

Y + XY

+ λXY = 0.
Assuming X(x)Y (y) is nonzero, dividing for X(x)Y (y) and subtract both
sides for
X

(x)
X(x)
, we ﬁnd:

X

(x)
X(x)
=
Y

(y)
Y (y)
+ λ.
The left hand side is a function of x while the right hand side is a function
of y. This says that they must equal to a constant. That is,

X

(x)
X(x)
=
Y

(y)
Y (y)
+ λ = δ.
where δ is a constant. This results in the following two ODEs
X

+ δX = 0 and Y

+ (λ −δ)Y = 0.
263
• If δ > 0 and λ −δ > 0 then
X(x) =Acos δx + Bsin δx
Y (y) =C cos (λ −δ)y + Dsin (λ −δ)y
• If δ > 0 and λ −δ < 0 then
X(x) =Acos δx + Bsin δx
Y (y) =Ce

−(λ−δ)y
+ De

−(λ−δ)y
• If δ = λ > 0 then
X(x) =Acos δx + Bsin δx
Y (y) =Cy + D
• If δ = λ < 0 then
X(x) =Ae

−δx
+ Be

−δx
Y (y) =Cy + D
• If δ < 0 and λ −δ > 0 then
X(x) =Ae

−δx
+ Be

−δx
Y (y) =C cos (λ −δ)y + Dsin (λ −δ)y
• If δ < 0 and λ −δ < 0 then
X(x) =Ae

−δx
+ Be

−δx
Y (y) =Ce

(λ−δ)y
+ De

(λ−δ)y
.
17.2 Let’s assume that the solution can be written in the form u(x, t) =
X(x)T(t). Substituting into the heat equation we obtain
X

X
=
T

kT
.
Since X only depends on x and T only depends on t, we must have that
there is a constant λ such that
X

X
= λ and
T

kT
= λ.
264 ANSWERS AND SOLUTIONS
This gives the two ordinary diﬀerential equations
X

−λX = 0 and T

−kλT = 0.
Next, we consider the three cases of the sign of λ.
Case 1: λ = 0
In this case, X

= 0 and T

= 0. Solving these equations we ﬁnd X(x) =
ax + b and T(t) = c.
Case 2: λ > 0
In this case, X(x) = Ae

λx
+ Be

λx
and T(t) = Ce
kλt
.
Case 3: λ < 0
In this case, X(x) = Acos

−λx + Bsin

−λx and and T(t) = Ce
kλt
.
17.3 r
2
R

(r) + rR

(r) −λR(r) = 0 and Θ

(θ) + λΘ(θ) = 0.
17.4 X

= (2 + λ)X, T

= λT, X(0) = 0, X(1) = 0.
17.5 X

−λX = 0, T

= kλT, X

(0) = 0 = X

(L).
17.6 u(x, t) = e
λ(x−t)
.
17.7 5X

−7X

−λX = 0 and 3Y

−λY

= 0.
17.8 u(x, y) = Ce
λx−
y
λ
.
17.9 u(x, y) = Ce
λx
y
λ
.
17.10 We look for a solution of the form u(x, y) = X(x)T(t). Substitut-
ing in the wave equation, we obtain
X

(x)T(t) −X(x)T

(t) = 0.
Assuming X(x)T(t) is nonzero, dividing for X(x)T(t) we ﬁnd:
X

(x)
X(x)
=
T

(t)
T(t)
.
265
The left hand side is a function of x while the right hand side is a function
of t. This says that they must equal to a constant. That is,
X

(x)
X(x)
=
T

(t)
T(t)
= λ
where λ is a constant. This results in the following two ODEs
X

−λX = 0 and T

−λT = 0.
The solutions of these equations depend on the sign of λ.
• If λ > 0 then the solutions are given
X(x) =Ae

λx
+ Be

λx
T(t) =Ce

λt
+ De

λt
where A, B, C, and D are constants. In this case,
u(x, t) = k
1
e
λ(x+t)
+ k
2
e
λ(x−t)
+ k
3
e
−λ(x+t)
+ k
4
e
−λ(x−t)
.
• If λ = 0 then
X(x) =Ax + B
T(t) =Ct + D
where A, B, and C are arbitrary constants. In this case,
u(x, t) = k
1
xt + k
2
x + k
3
t + k
4
.
• If λ < 0 then
X(x) =Acos

−λx + Bsin

−λx
T(t) =Acos

−λt + Bsin

−λt
where A, B, C, and D are arbitrary constants. In this case,
u(x, t) =k
1
cos

−λx cos

−λt + k
2
cos

−λx sin

−λt
+k
3
sin

−λx cos

−λt + k
4
sin

−λx sin

−λt.
17.11 (a) u(r, t) = R(r)T(t), T

(t) = −kλT,
1
r
(rR

)

= −λR.
266 ANSWERS AND SOLUTIONS
(b) u(x, t) = X(x)T(t), T

+ αT = −kλT, R

= −λR.
(c) u(x, t) = X(x)T(t), T

= −λT, kX

−aX

= −λX.
(d) u(x, t) = X(x)Y (y), X

= −λX, Y

= −λY.
(e) u(x, t) = X(x)T(t), T

= −kλT, X

= −λX.
17.12 u(x, y) = Ce
λ(x+y)
.
17.13 X

= λX, Y

−Y

+ Y = λY.
Section 18
18.1 u(x, t) = sin
_
π
2
x
_
e

π
2
k
4
t
+ 3 sin
_

2
x
_
e

25π
2
k
4
t
.
18.2 u(x, t) =
8d
π
3

n=1
1
(2n−1)
3
sin
_
(2n−1)π
L
x
_
e

k(2n−1)
2
π
2
L
2
t
.
18.3 u(x, t) =
2
π

4
π

n=1
1
(4n
2
−1)
cos
_
2nπ
L
x
_
e
−k
4n
2
π
2
L
2
t
.
18.4 u(x, t) =

n=1
C
n
sin
_

L
x
_
e

n
2
π
2
L
2
t
where
C
n
=
_
_
_

4

n = 2, 6, 10, · · ·
0 n = 4, 8, 12, · · ·
6

n is odd.
18.5 u(x, t) = 6 sin
_

L
x
_
e
−81π
2
L
2
t
.
18.6 u(x, t) =
1
2
+

n=1
C
n
cos
_

L
x
_
e

n
2
π
2
L
2
t
where
C
n
=
_
_
_

2

n = 1, 5, 9, · · ·
2

n = 3, 7, 11, · · ·
0 n is even
18.7 u(x, t) = 6 + 4 cos
_

L
x
_
e

2
L
2
t
.
18.8 u(x, t) = −3 cos
_

L
x
_
e

64π
2
L
2
t
.
18.9
u(x, t) =

n=0
a
n
cos
_
(2n + 1)π
2L
x
_
e

(2n+1)
2
π
2
4L
2
t
.
267
As t →∞, e

(2n+1)
2
π
2
4L
2
t
→0 for each n ∈ N. Hence, u(x, t) →0.
18.10
u(x, t) =

n=0
a
n
cos
_
(nπ
L
x
_
e

1+
n
2
π
2
L
2

t
.
As t →∞, e

1+
n
2
π
2
L
2

t
→0 for each n ∈ N. Hence, u(x, t) →0.
18.11 (b) We have
E

(t) =2
_
1
0
w(x, t)w
t
(x, t)dx
=2
_
1
0
w(x, t)[w
xx
(x, t) −w(x, t)]dx
= 2w(x, t)w
x
(x, t)|
1
0
−2
__
1
0
w
2
x
(x, t)dx +
_
1
0
w
2
(x, t)dx
_
=−2
__
1
0
w
2
x
(x, t)dx +
_
1
0
w
2
(x, t)dx
_
≤ 0
Hence, E is decreasing, and 0 ≤ E(t) ≤ E(0) for all t > 0.
(c) Since w(x, 0) = 0, we must have E(0) = 0. Hence, E(t) = 0 for all t ≥ 0.
This implies that w(x, t) = 0 for all t > 0 and all 0 < x < 1. Therefore
u
1
(x, t) = u
2
(x, t). This means that the given problem has a unique solution.
18.12 (a) u(0, t) = 0 and u
x
(1, t) = 0.
(b) Let’s assume that the solution can be written in the form u(x, t) =
X(x)T(t). Substituting into the heat equation we obtain
X

X
=
T

kT
.
Since X only depends on x and T only depends on t, we must have that
there is a constant λ such that
X

X
= λ and
T

kT
= λ.
This gives the two ordinary diﬀerential equations
X

−λX = 0 and T

−kλT = 0.
268 ANSWERS AND SOLUTIONS
As far as the boundary conditions, we have
u(0, t) = 0 = X(0)T(t) =⇒X(0) = 0
and
u
x
(1, t) = 0 = X

(1)T(t) =⇒X

(1) = 0.
Note that T is not the zero function for otherwise u ≡ 0 and this contradicts
our assumption that u is the non-trivial solution.
(c) We have X

=

−λcos

−λx and X

= λsin

−λx. Thus, X

−λX = 0.
Moreover X(0) = 0. Now, X

(1) = 0 implies cos

−λ = 0 or

−λ =
_
n −
1
2
_
π, n ∈ N. Hence, λ = −
_
n −
1
2
_
2
π
2
.
18.13 (a) Let’s assume that the solution can be written in the form u(x, t) =
X(x)T(t). Substituting into the heat equation we obtain
X

X
=
T

kT
.
Since the LHS only depends on x and the RHS only depends on t, then there
must be a constant λ such that
X

X
= λ and
T

kT
= λ.
This gives the two ordinary diﬀerential equations
X

−λX = 0 and T

−kλT = 0.
As far as the boundary conditions, we have
u(0, t) = 0 = X(0)T(t) =⇒X(0) = 0
and
u(L, t) = 0 = X(L)T(t) =⇒X(L) = 0.
Note that T is not the zero function for otherwise u ≡ 0 and this contradicts
our assumption that u is the non-trivial solution.
Next, we consider the three cases of the sign of λ.
Case 1: λ = 0
In this case, X

= 0. Solving this equation we ﬁnd X(x) = ax + b. Since
X(0) = 0 we ﬁnd b = 0. Since X(L) = 0 we ﬁnd a = 0. Hence, X ≡ 0 and
269
u(x, t) ≡ 0. That is, u is the trivial solution.
Case 2: λ > 0
In this case, X(x) = Ae

λx
+Be

λx
. Again, the conditions X(0) = X(L) =
0 imply A = B = 0 and hence the solution is the trivial solution.
Case 3: λ < 0
In this case, X(x) = Acos

−λx + Bsin

−λx. The condition X(0) = 0
implies A = 0. The condition X(L) = 0 implies Bsin

−λL = 0. We must
have B = 0 otherwise X(x) = 0 and this leads to the trivial solution. Since
B = 0, we obtain sin

−λL = 0 or

−λL = nπ where n ∈ N. Solving for λ
we ﬁnd λ = −
n
2
π
2
L
2
. Thus, we obtain inﬁnitely many solutions given by
X
n
(x) = A
n
sin

L
x, n ∈ N.
Now, solving the equation
T

−λkT = 0
by the method of separation of variables we obtain
T
n
(t) = B
n
e

n
2
π
2
L
2
kt
, n ∈ N.
Hence, the functions
u
n
(x, t) = C
n
sin
_

L
x
_
e

n
2
π
2
L
2
kt
, n ∈ N
satisfy u
t
= ku
xx
and the boundary conditions u(0, t) = u(L, t) = 0.
Now, in order for these solutions to satisfy the initial value condition u(x, 0) =
6 sin
_
9πx
L
_
, we invoke the superposition principle of linear PDE to write
u(x, t) =

n=1
C
n
sin
_

L
x
_
e

n
2
π
2
L
2
kt
. (1)
To determine the unknown constants C
n
we use the initial condition u(x, 0) =
6 sin
_
9πx
L
_
in (1) to obtain
6 sin
_
9πx
L
_
=

n=1
C
n
sin
_

L
x
_
.
270 ANSWERS AND SOLUTIONS
By equating coeﬃcients we ﬁnd C
9
= 6 and C
n
= 0 if n = 9. Hence, the
solution to the problem is given by
u(x, t)6 sin
_
9πx
L
_
e

81π
2
L
2
kt
.
(b) Similar to (a), we ﬁnd
u(x, t) = 3 sin
_
π
L
x
_
e

π
2
kt
L
2
−sin
_

L
x
_
e

2
kt
L
2
18.14 u(x, t) = cos
_
πx
L
_
e

pi
2
kt
L
2
+ 4 cos
_
5πx
L
_
e

25pi
2
kt
L
2
.
(b) u(x, t) = 5.
18.15 u(x, t) = 6 sin xe
−8t
.
Section 19
19.1 u(x, y) =

n=1
B
n
sin
_

b
y
_
sinh
_

b
x
_
where
B
n
=
_
2
b
_
b
0
f
2
(y) sin
_

b
y
_
dy
_
_
sinh
_

b
a
__
−1
.
19.2 u(x, y) =

n=1
B
n
sin

a
x sinh
_

a
(y −b)
_
where
B
n
=
_
2
a
_
a
0
g
1
(x) sin
_

a
x
_
dx
_
[sinh
_

a
b
_
]
−1
.
19.3 u(x, y) = 2xy +
3
sinh π
sin πx sinh πy.
19.4 If u(x, y) = x
2
− y
2
then u
xx
= 2 and u
yy
= −2 so that ∆u = 0.
If u(x, y) = 2xy then u
xx
= u
yy
= 0 so that ∆u = 0.
19.5
u(x, y) =

n=1
[A
n
cosh
_

L
y
_
+ B
n
sinh
_

L
y
_
] sin

L
x.
where
A
n
=
_
2
L
_
L
0
(f
1
(x) + f
2
(x)) sin

L
xdx
_ _
cosh
_
nπH
2L
__
−1
271
and
B
n
=
_
2
L
_
L
0
(f
2
(x) −f
1
(x)) sin

L
xdx
_ _
sinh
_
nπH
2L
__
−1
19.6 (a) Diﬀerentiating term by term with respect to x we ﬁnd
u
x
+ iv
x
=

n=0
na
n
(x + iy)
n−1
.
Likewise, diﬀerentiating term by term with respect to y we ﬁnd
u
y
+ iv
y
=

n=0
na
n
i(x + iy)
n−1
.
Multiply this equation by i we ﬁnd
−iu
y
+ v
y
=

n=0
na
n
(x + iy)
n−1
.
Hence, u
x
+ iv
x
= v
y
−iu
y
which implies u
x
= v
y
and v
x
= −u
y
.
(b) We have u
xx
= (v
y
)
x
= (v
x
)
y
= −u
yy
so that ∆u = 0. Similar argument
for ∆v = 0.
19.7 Polar and Cartesian coordinates are related by the expressions x =
r cos θ and y = r sin θ where r = (x
2
+ y
2
)
1
2
and tan θ =
y
x
. Using the chain
272 ANSWERS AND SOLUTIONS
rule we obtain
u
x
=u
r
r
x
+ u
θ
θ
x
= cos θu
r

sin θ
r
u
θ
u
xx
=u
xr
r
x
+ u

θ
x
=
_
cos θu
rr
+
sin θ
r
2
u
θ

sin θ
r
u

_
cos θ
+
_
−sin θu
r
+ cos θu

cos θ
r
u
θ

sin θ
r
u
θθ
__

sin θ
r
_
u
y
=u
r
r
y
+ u
θ
θ
y
= sin θu
r
+
cos θ
r
u
θ
u
yy
=u
yr
r
y
+ u

θ
y
=
_
sin θu
rr

cos θ
r
2
u
θ
+
cos θ
r
u

_
sin θ
+
_
cos θu
r
+ sin θu

sin θ
r
u
θ
+
cos θ
r
u
θθ
__
cos θ
r
_
Substituting these equations into (21.1) we obtain the dersired equation.
19.8 u(x, y) = u
1
(x, y) + u
2
(x, y) + u
3
(x, y) + u
4
(x, y) where
u
1
(x, y) = 0
u
2
(x, y) =

n=1
_

2

·
(−1)
n
sinh
_
3nπ
2
_
_
sin

2
x sinh
_

2
y
_
u
3
(x, y) =
1
sinh
_

3
_ sinh
_
4π(x −2)
3
_
sin
_

3
y
_
u
4
(x, y) =

n=1
14(1 −(−1)
n
)
nπ sinh
_
2nπ
3
_ sin
_

3
y
_
sinh
_

3
x
_
.
19.9
u(x, y) =
4
sinh
_
πL
2H
_
_
sinh
_
πx
2H
_
−sinh
_
π(x −L)
2H
__
cos
πy
2H
.
273
19.10 u(x, t) = A
0
+

n=1
A
n
e

λnx
cos

λ
n
y where
A
0
=
1
H
_
H
0
f(y)dy
A
n
=
2
H
_
H
0
f(y) cos

H
ydy.
19.11 u(x, y) =
20
Y
1
(H)
Y
1
(y) sin
_
πx
L
_

5
Y
3
(H)
sin
_
3πx
L
_
.
19.12 u(x, y) = sin (2πx)e
−2πy
.
19.13 u(x, y) = y.
19.14 u(x, y) =
1
2
x
2

1
2
y
2
−ax
b
y + C where C is an arbitrary constant.
19.15 u(x, y) =
2 cosh 3y sin 3x
cosh 6

5 cosh 10y sin 10x
cosh 20
.
Section 20
20.1 u(r, θ) = 3r
5
sin 5θ.
20.2 u(r, θ) =
π
4
+

n=1
r
n
_
1−(−1)
n
n
2
π
cos nθ +
sin nθ
n
_
.
20.3 u(r, θ) = C
0
+ r
2
cos 2θ.
20.4 Substituting C
0
, A
n
, and B
n
into the right-hand side of u(r, θ) we ﬁnd
u(r, θ) =
1

_

0
f(φ)dφ +

n=1
r
n
πa
n
_

0
f(φ) [cos nφcos nθ + sin nφsin nθ] dφ
=
1

_

0
f(φ)
_
1 + 2

n=1
_
r
a
_
n
cos n(θ −φ)
_
dφ.
20.5 (a) We have e
it
= cos t + i sin t and e
−it
= cos t − i sin t. The result
follows by adding these two equalities and dividing by 2.
(b) This follows from the fact that
cos n(θ −φ) =
1
2
(e
in(θ−φ)
+ e
−in(θ−φ)
).
274 ANSWERS AND SOLUTIONS
(c) We have |q
1
| =
r
a
_
cos (θ −φ)
2
+ sin (θ −φ)
2
=
r
a
< 1 since 0 < r < a. A
similar argument shows that |q
2
| < 1.
20.6 (a) The ﬁrst sum is a convergent geometric series with ratio q
1
and
sum

n=1
_
r
a
_
n
e
in(θ−φ)
=
r
a
e
i(θ−φ)
1 −q
1
=
re
i(θ−φ)
a −re
i(θ−φ)
Similar argument for the second sum.
(b) We have
1 + 2

n=1
_
r
a
_
n
cos n(θ −φ) =1 +
re
i(θ−φ)
a −re
i(θ−φ)
+
re
−i(θ−φ)
a −re
−i(θ−φ)
=1 +
r
ae
−i(θ−φ)
−r
+
r
ae
−i(θ−φ)
−r
=1 +
r
a cos (θ −φ) −r −ai sin (θ −φ)
+
r
a cos (θ −φ) −r + ai sin (θ −φ)
=1 +
r[a cos (θ −φ) −r + ai sin (θ −φ)]
a
2
+ 2ar cos (θ −φ) + r
2
+
r[a cos (θ −φ) −r −ai sin (θ −φ)]
a
2
−2ar cos (θ −φ) + r
2
=
a
2
−r
2
a
2
−2ar cos (θ −φ) + r
2
.
275
20.7 We have
u(r, θ) =
1

_

0
f(φ)
_
1 + 2

n=1
_
r
a
_
n
cos n(θ −φ)
_

=
1

_

0
f(φ)
a
2
−r
2
a
2
−2ar cos (θ −φ) + r
2

=
a
2
−r
2

_

0
f(φ)
a
2
−2ar cos (θ −φ) + r
2
dφ.
20.8 u(r, θ) = 2

n=1
(−1)
n+1
r
n sin nθ
n
.
20.9 (a) Diﬀerentiating u(r, t) = R(r)T(t) with respect to r and t we ﬁnd
u
tt
= RT

and u
r
= R

T and u
rr
= R

T.
Substituting these into the given PDE we ﬁnd
RT

= c
2
_
R

T +
1
r
R

T
_
Dividing both sides by c
2
RT we ﬁnd
1
c
2
T

T
=
R

R
+
1
r
R

R
.
Since the RHS of the above equation depends on r only, and the LHS depends
on t only, they must equal to a constant λ.
(b) The given boundary conditions imply
u(a, t) = 0 = R(a)T(t) =⇒R(a) = 0
u(r, 0) = f(r) = R(r)T(0)
u
t
(r, 0) = g(r) = R(r)T

(0).
If λ = 0 then R

+
1
r
R

= 0 and this implies R(r) = C ln r. Using the condition
R(a) = 0 we ﬁnd C = 0 so that R(r) = 0 and hence u ≡ 0. If λ > 0 then
T

−λc
2
T = 0. This equation has the solution
T(t) = Acos (c

λt) + Bsin (c

λt).
276 ANSWERS AND SOLUTIONS
The condition u(r, 0) = f(r) implies that A = f(r) which is not possible.
Hence, λ < 0.
20.10 (a) Follows from the ﬁgure and the deﬁnitions of trigonometric func-
tions in a right triangle.
(b) The result follows from equation (20.1).
20.11 By the maximum principle we have
min
(x,y)∈∂Ω
u(x, y) ≤ u(x, y) ≤ max
(x,y)∈∂Ω
u(x, y), ∀(x, y) ∈ Ω
But min
(x,y)∈∂Ω
u(x, y) = u(1, 0) = 1 and max
(x,y)∈∂Ω
u(x, y) = u(−1, 0) = 3.
Hence,
1 ≤ u(x, y) ≤ 3
and this implies that u(x, y) > 0 for all (x, y) ∈ Ω.
20.12 (i) u(1, 0) = 4 (ii) u(−1, 0) = −2.
20.13 Using the maximum principle and the hypothesis on g
1
and g
2
, for
all (x, y) ∈ Ω ∪ ∂Ω we have
min
(x,y)∈∂Ω
u
1
(x, y) = min
(x,y)∈∂Ω
g
1
(x, y)
≤u
1
(x, y) ≤ max
(x,y)∈∂Ω
u
1
(x, y)
= max
(x,y)∈∂Ω
g
1
(x, y) < max
(x,y)∈∂Ω
g
2
(x, y)
≤ min
(x,y)∈∂Ω
g
1
(x, y) = min
(x,y)∈∂Ω
u
2
(x, y)
≤u
2
(x, y) ≤ max
(x,y)∈∂Ω
u
2
(x, y) = max
(x,y)∈∂Ω
g
2
(x, y).
20.14 We have
∆(r
n
cos (nθ)) =

2
∂r
2
(r
n
cos (nθ)) +
1
r

∂r
(r
n
cos (nθ)) +
1
r
2

2
∂θ
2
(r
n
cos (nθ))
=n(n −1)r
n−2
cos (nθ) + nr
n−2
cos (nθ) −r
n−2
n
2
cos (nθ) = 0
Likewise, ∆(r
n
sin (nθ)) = 0.
277
20.15 u(r, θ) =
1
2

r
2
2a
2
cos 2θ.
20.16 u(r, θ) = ln 2 + 4
_
a
r
_
3
cos 3θ.
Section 21
21.1 Convergent.
21.2 Divergent.
21.3 Convergent.
21.4
1
s−3
, s > 3.
21.5
1
s
2

5
s
, s > 0.
21.6 f(t) = e
(t−1)
2
does not have a Laplace transform.
21.7
4
s

4
s
2
+
2
s
3
, s > 0.
21.8
e
−s
s
2
, s > 0.
21.9 −
e
−2s
s
+
1
s
2
(e
−s
−e
−2s
), s = 0.
21.10 −
t
n
e
−st
s
+
n
s
_
t
n−1
e
−st
dt, s > 0.
21.11 (a) 0 (b) 0.
21.12
5
s+7
+
1
s
2
+
2
s−2
, s > 2.
21.13 3e
2t
, t ≥ 0.
21.14 −2t + e
−t
, t ≥ 0.
21.15 2(e
−2t
+ e
2t
), t ≥ 0.
21.16
2
s−1
+
5
s
, s > 1.
278 ANSWERS AND SOLUTIONS
21.17
e
−s
s−3
, s > 3.
21.18
1
2
_
1
s

s
2
s
2
+4ω
2
_
, s > 0.
21.19
3
s
2
+26
, s > 0.
21.20
s−3
(s−3)
2
+9
, s > 3.
21.21
2
(s−4)
3
+
3
(s−4)
2
+
5
s−4
, s > 4.
21.22 2 sin 5t + 4e
3t
, t ≥ 0.
21.23
5
6
e
3t
t
3
, t ≥ 0.
21.24
_
0, 0 ≤ t < 2
e
9(t−2)
, t ≥ 2.
21.25 3e
3t
−3e
−t
, t ≥ 0.
21.26 4[e
3(t−5)
−e
−3(t−5)
]H(t −5), t ≥ 0.
21.27 y(t) = 2e
−4t
+3[H(t−1)−H(t−3)] −3[e
−4(t−1)
H(t−1)−e
−4(t−3)
H(t−
3)], t ≥ 0.
21.28
1
5
e
3t
+
1
20
e
−2t

1
4
e
2t
, t ≥ 0.
21.29
e
t
−e
−2t
3
.
21.30
t
2
sin t.
21.31
t
5
120
.
21.32
1
2
−e
−t
+
1
2
e
−2t
.
21.33 −t +
e
t
2

e
−t
2
.
Section 22
279
22.1 u(x, t) = sin (x −t) −H(t −x) sin (x −t).
22.2 u(x, t) = [sin (x −t) −H(t −x) sin (x −t)]e
−t
.
22.3 u(x, t) = 2e
−4π
2
t
sin πx + 6e
−16π
2
t
sin 2πx.
22.4 u(x, t) = [sin (x −t) −H(t −x) sin (x −t)]e
t
.
22.5 u(x, t) = t
2
e
−x
−te
−x
+ t.
22.6 u(x, t) =
_
t −
1
2
x
2
_
H
_
t −
1
2
x
2
_
.
22.7 u(x, t) = L
−1
_
e

s
c
x
s
2
+1
_
= H
_
t −
x
c
_
sin
_
t −
x
c
_
.
22.8 u(x, t) = 2 sin x cos 3t.
22.9 u(x, y) = y(x + 1) + 1.
22.10 u(x, t) = L
−1
_
e

s
c
x
s
2
+1
_
= h
_
t −
x
c
_
sin
_
t −
x
c
_
.
22.11 u(x, t) = e
−5x
e
−4t
H(t).
22.12 u(x, t) = L
−1
_

T
s
e

s
c
x
+
T
s
_
.
22.13 u(x, t) = 5e
−3π
2
t
sin (πx).
22.14 u(x, t) = 40e
−t
cos
x
2
.
22.15 u(x, t) = 3 sin πx cos 2πt.
Section 23
23.1
(−1)
n
i

.
23.2 f(x) =
1
2
+

n=1
1

sin
_

2
_
(e
inx
+ e
−inx
).
280 ANSWERS AND SOLUTIONS
23.3 f(x) =
sinh aπ
π

n=−∞
(−1)
n
(a+in)
(a
2
+n
2
)
e
inx
.
23.4 f(x) =
e
ix
−e
−ix
2i
.
23.5 f(x) =
1

_
T +

−1
n=−∞
i
n
[e
−int
−1]e
int
+

n=1
i
n
[e
−int
−1]e
int
_
.
23.6 (a) f(x) =
π
2
3
+

−1
n=−∞
2
n
2
(−1)
n
e
inx
+

n=1
2
n
2
(−1)
n
e
inx
.
(b) f(x) =
π
2
3
+

n=1
4
n
2
(−1)
n
cos nx.
23.7 (a)
a
0
=2
_ 1
2

1
2
sin nπxdx = −
2
π
[cos
π
2
−cos −
π
2
] = 0
a
n
=2
_ 1
2

1
2
sin nπx cos 2nπxdx = 0
b
n
=2
_ 1
2

1
2
sin nπx sin 2nπxdx =
8(−1)
n
n
π −4n
2
π
(b) f(x) =
4
π

n=−∞
(−1)
n
n
i(1−4n
2
)
e
2nπix
.
23.8 (a)
a
0
=
1
2
_
2
−2
(2 −x)dx = 4
a
n
=
1
2
_
2
−2
(2 −x) cos
_

2
x
_
dx = 0
b
n
=
1
2
_
2
−2
(2 −x) sin
_

2
x
_
dx =
4(−1)
n

(b) f(x) = 2 +

−1
n=−∞
2(−1)
n+1
i

e
(
inπ
2
x)
+

n=1
2(−1)
n+1
i

e
(
inπ
2
x)
.
23.9 a
n
= c
n
+ c
−n
= 0. We have for |n| odd b
n
= i
4
inπ
=
4

and for
|n| even b
n
= 0.
23.10 Note that for any complex number z we have z + z = 2Re(z) and
z −z = −2iRe(z). Thus,
c
n
+ c
n
= a
n
281
which means that a
n
= 2Re(c
n
). Likewise, we have
c
n
−c
n
= ib
n
That is ib
n
= −2iIm(c
n
). Hence, b
n
= −2Im(c
n
).
23.11 a
n
= 2Re(c
n
) =
1
πn
sin (nT) and b
n
=
1−cos (nT)

.
23.12 f(x) = i

n=−∞
i sin (2−inπ)
2−inπ
e
inπ
2
x
.
23.13 (a) We have
f(t) =
_
1 0 < t < 1
0 1 < t < 2
and f(t + 2) = f(t) for all t ∈ R.
(b) We have
a
0
=
2
L
_
L
0
f(x)dx =
_
2
0
dx =
_
1
0
dx = 1
a
n
=
_
1
0
cos nπxdx =
sin nπ

= 0.
(c) We have
b
n
=
_
1
0
sin nπxdx =
1 −cos nπ

=
1 −(−1)
n

.
Hence,
b
n
=
_
2

if n is odd
0 if n is even
(d) We have c
0
=
a
0
2
=
1
2
and for n ∈ N we have
c
n
=
a
n
−ib
n
2
=
_

i

if n is odd
0 if n is even
23.14 sin 3x =
1
2
(e
3ix
−e
−3ix
).
23.15 e
−ins
_
1−e
−inh
2πin
_
.
282 ANSWERS AND SOLUTIONS
Section 24
24.1
ˆ
f(ξ) =
_
2
sin ξ
ξ
if ξ = 0
2 if ξ = 0.
24.2
∂ˆ u
∂t
+ iξcˆ u = 0
ˆ u(ξ, 0) =
ˆ
f(ξ).
24.3

2
ˆ u
∂t
2
= −c
2
ξ
2
ˆ u
ˆ u(ξ, 0) =
ˆ
f(ξ)
ˆ u
t
(ξ, 0) = ˆ g(ξ).
24.4
ˆ u
yy
= ξ
2
ˆ u
ˆ u(ξ, 0) = 0, ˆ u(ξ, L) =
2 sin ξa
ξ
.
24.5
1
α−iξ
+
1
α+iξ
=

α
2

2
.
24.6 We have
F[e
−x
H(x)] =
_

−∞
e
−x
H(x)e
−iξx
dx
=
_

0
e
−x(1+iξ)
dx = −
e
−x(1+iξ)
1 + iξ
¸
¸
¸
¸

0
=
1
1 + iξ
.
24.7 Using the duality property, we have
F
_
1
1 + ix
_
= F[F[e
−ξ
H(ξ)]] = 2πe
ξ
H(−ξ).
283
24.8 We have
F[f(x −α)] =
_

−∞
f(x −α)e
−iξx
dx
=e
−iξα
_

−∞
f(u)e
−iξu
du
=e
−iξα
ˆ
f(ξ)
where u = x −α.
24.9 We have
F[e
iαx
f(x)] =
_

−∞
e
iαx
f(x)e
−iξx
dx =
_

−∞
e
ix(α−ξ
f(x)e
−iξx
dx =
ˆ
f(ξ −α).
24.10 We will just prove the ﬁrst one. We have
F[cos (αx)f(x)] =F[
f(x)e
iαx
2
+ f(x)
e
−iαx
2
=
1
2
[F[f(x)e
iαx
] +F[f(x)e
−iαx
]]
=
1
2
[
ˆ
f(x −α) +
ˆ
f(x + α)].
24.11 Using the deﬁnition and integration by parts we ﬁnd
F[f

(x)] =
_

−∞
f

(x)e
−iξx
dx
= f(x)e
−iξx
¸
¸

−∞
+ (iξ)
_

−∞
f(x)e
−iξx
dx
=f(x) cos ξx −if(x) sin ξx + (iξ)
ˆ
f(ξ) = (iξ)
ˆ
f(ξ)
where we used the fact that lim
x→∞
f(x) = 0.
24.12
2
ξ
2
(1 −cos ξ).
24.13
2

(1 −cos ξa).
284 ANSWERS AND SOLUTIONS
24.14 F
−1
[
ˆ
f(ξ)] =
1

e

x
2
2
.
24.15 F
−1
_
1
a+iξ
_
= e
−ax
, x ≥ 0.
Section 25
25.1 u(x, t) = f(x) ∗ F
−1
[−
1
|ξ|
e
−|ξ|y
].
25.2 u(x, t) = F
−1
[u(ξ, t)] = e

(x−ct)
2
4
.
25.3
u(x, t) =
_
γ

e
−αt
F
−1
[e
−ξ
2
(kt+
γ
4
)
]
=
_
γ

e
−αt
·
_
π
kt + γ/4
· e

x
2
4(kt+γ/4)
=

γ4kt + γe

x
2
4kt+γ
e
−αt
.
25.4 u(x, t) =
1

4πkt
_

0
e

(x−s)
2
4kt
ds.
25.5
u(x, t) =e
t
F
−1
[e
−ξ
2
t
]
=e
−αt
1

4πt
e

x
2
4t
.
25.6 We have
_

−∞
e
−|ξ|y
e
iξx
dξ =
_
0
−∞
e
ξy
e
iξx
dξ +
_

0
e
−ξy
e
iξx

=
1
y + ix
e
ξ(y+ix)
¸
¸
¸
¸
0
−∞
+
1
−y + ix
e
ξ(−y+ix)
¸
¸
¸
¸

0
=
1
y + ix
+
1
−y + ix
=
2y
x
2
+ y
2
.
285
25.7
u(x, y) =
1

_

−∞
ˆ
f(ξ)e
−|ξ|y
e
iξx

=
1

f(x) ∗
_
2y
x
2
+ y
2
_
=
1

_

−∞
f(x)
2y
(x −ξ)
2
+ y
2
dξ.
25.8 ˆ u
tt
+ (α + β)ˆ u
t
+ αβˆ u = −c
2
ξ
2
ˆ u.
25.9 u(x, t) = e
−(x−3t)
.
25.10 u(x, t) = e
−(x−kt)
.
25.11 u(x, t) =
1

4πkt
_

−∞
e
−s
2

(x−s)
2
4kt
ds.
25.12 u(x, t) = (x −ct)
2
.
25.13 u(x, t) = f(x) ∗ F
−1
[−
1
|ξ|
e
−|ξ|y
].

2

Preface
Partial diﬀerential equations are often used to construct models of the most basic theories underlying physics and engineering. The goal of this book is to develop the most basic ideas from the theory of partial diﬀerential equations, and apply them to the simplest models arising from the above mentioned ﬁelds. It is not easy to master the theory of partial diﬀerential equations. Unlike the theory of ordinary diﬀerential equations, which relies on the fundamental existence and uniqueness theorem, there is no single theorem which is central to the subject. Instead, there are separate theories used for each of the major types of partial diﬀerential equations that commonly arise. It is worth pointing out that the preponderance of diﬀerential equations arising in applications, in science, in engineering, and within mathematics itself, are of either ﬁrst or second order, with the latter being by far the most prevalent. We will mainly cover these two classes of PDEs. This book is intended for a ﬁrst course in partial diﬀerential equations at the advanced undergraduate level for students in engineering and physical sciences. It is assumed that the student has had the standard three semester calculus sequence, and a course in ordinary diﬀerential equations.

Marcel B Finan August 2009

3

4

PREFACE

Contents
Preface 3

Preliminaries 7 1 Some Results of Calculus . . . . . . . . . . . . . . . . . . . . . . . 7 2 Sequences of Functions: Pointwise and Uniform Convergence . . . 14 Review of Some ODEs Results 3 The Method of Integrating Factor . . . . . . . . . . . . . . . . . . 4 The Method of Separation of Variables for ODEs . . . . . . . . . 5 Second Order Linear ODEs . . . . . . . . . . . . . . . . . . . . . 23 23 28 33

Introduction to PDEs 43 6 The Basic Concepts . . . . . . . . . . . . . . . . . . . . . . . . . 43 7 Solutions and Related Topics . . . . . . . . . . . . . . . . . . . . 52 First Order Partial Diﬀerential Equations 8 Classiﬁcation of First Order PDEs . . . . . . . . . . . . . . 9 The One Dimensional Spatial Transport Equations . . . . . 10 The Method of Characteristics . . . . . . . . . . . . . . . . 11 The Cauchy Problem for First Order Quasilinear Equations 65 65 71 80 87 101 101 106 114 123 134 142

. . . .

. . . .

. . . .

Second Order Linear Partial Diﬀerential Equations 12 Second Order PDEs in Two Variables . . . . . . . . . . . . . . 13 Hyperbolic Type: The Wave equation . . . . . . . . . . . . . . 14 Parabolic Type: The Heat Equation in One-Dimensional Space 15 An Introduction to Fourier Series . . . . . . . . . . . . . . . . 16 Fourier Sines Series and Fourier Cosines Series . . . . . . . . . 17 Separation of Variables for PDEs . . . . . . . . . . . . . . . . 5

. . . . . .

. . . . . . 157 20 Laplace’s Equations in Circular Regions . . . . . . . . . . 181 22 Solving PDEs Using Laplace Transform . . . . . . . . . . . . . . . . . . . . . . 196 The Fourier Transform Solutions for PDEs 203 23 Complex Version of Fourier Series . . . . . . . . . 217 Answers and Solutions 225 . . . . . . . . . . . . . . . . . . . . . . . . . 168 The Laplace Transform Solutions for PDEs 181 21 Essentials of the Laplace Transform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 19 Elliptic Type: Laplace’s Equations in Rectangular Domains . 209 25 Applications of Fourier Transforms to PDEs . . . . . .6 CONTENTS 18 Solutions of the Heat Equation by the Separation of Variables Method . . . . . . . . 203 24 Two Dimensional Fourier Transforms . . .

b (c) Suppose that f : [a. b] → R is continuous and non-negative.Preliminaries In this chapter we include some results from calculus which we will use often in the study of partial diﬀerential equations.1 b (a) Suppose that f is continuous on an interval I ⊂ R such that a f (x)dx = 0 for all subintervals [a. That is. b]. b]. Hence. suppose that x0 ∈ [a. b] such that f (x0 ) > 0. a Since x was arbitrary. By the Fundamental Theorem of Calculus we have 0= d dx x f (t)dt = f (x). If a f (x)dx = 0 then f (x) = 0 on [a. Then f (x) = 0 on [a. (b) Suppose the contrary. Then f (x) = 0 for all x ∈ I. Details and proof of these results can be found in most calculus books. The δ deﬁnition of continuity of f at x0 guarantees the existence of an open interval I ⊂ [a. because f (x) ≥ 0 we must have b f (x)dx ≥ a I f (x)dx > 0 7 . b] → R is continuous such that a f (x)g(x)dx = 0 for all continuous functions g on [a. Solution. (a) Fix a ∈ I. Theorem 1. we have f (x) = 0 for all x ∈ I. Let x ∈ I. b (b) Suppose that f : [a. b] ⊂ I. b] centered at x0 such that f (x) > 0 for all x ∈ I. 1 Some Results of Calculus The ﬁrst result provides a mean of showing when a function is zero on an interval. b].

The second. b] ⊂ J and [c. t) then ∂u ∂u ∂x ∂u ∂y = + . The ﬁrst is the Leibnitz notation that employs the symbol ∂ to denote partial derivative. 2 An important formula of diﬀerentiation is the so-called chain rule. g : [a. d] ⊂ I then f (x. y) is deﬁned for x in an interval I and y in an interval J such that b a c d f (x. Apply part (b) of previous theorem to the function h(x) = g(x) − f (x) Partial Derivatives For multivariable functions. We conclude that f (x) = 0 on [a. b]. b] → ∞ be such that f (x) ≤ g(x) for all x in [a. if f (x. y)dxdy = 0 for all [a. y) where x = x(s. a more compact notation. u = u(x. t) and y = y(s.1 The above theorem remains valid for functions in two variables. while uxx represents ∂ u .1 Let f. ut represents ∂u . is to use subscripts to indicate partial 2 derivatives. For example. Example 1. and uxxt ∂t ∂x2 3u becomes ∂∂ x∂t . b].8 PRELIMINARIES which contradicts our assumption that the integral is zero. there are two common notations for partial derivatives. ∂s ∂x ∂s ∂y ∂s Likewise. Show that if b (g(x) − f (x))dx = 0 then f (x) ≡ g(x) on [a. ∂u ∂u ∂x ∂u ∂y = + . a Solution. (c) This follows from (b) by taking g(x) = f (x) Remark 1. y) = 0 over the rectangle I × J. b]. and we shall employ them interchangeably. ∂t ∂x ∂t ∂y ∂t If . For example.

where x = tes and y = s + t. Assume that ∂f is continuous on this rectangle. t)b (t) − f (a(t). Theorem 1.2 Compute the partial derivatives indicated: ∂ (a) ∂y (y 2 sin xy) (b) ∂2 [ex+y ]2 ∂x2 9 Solution. Deﬁne the function ∂t b(t) J(t) = a(t) f (x. ∂x2 [ex+y ]2 = ∂x 2e2(x+y) = ∂x Example 1. Solution.1 SOME RESULTS OF CALCULUS Example 1. Let f (x. t)dx ∂t . y) = sin (x2 + y 2 ). ut =ux xt + uy yt = 2x cos (x2 + y 2 )es + 2y cos (x2 + y 2 ) =[2te2s + 2(s + t)] cos [t2 e2s + (s + t)2 ] Often we must diﬀerentiate an integral with respect to a parameter which may appear in the limits of integration. t)dx where a(t) and b(t) are continuously diﬀerentiable functions of t such that a ≤ a(t) ≤ b(t) ≤ b. t)a (t) + a(t) ∂f (x. Find us and ut .3 Suppose u(x. Thus.2 dJ d = dt dt b(t) f (x. or in the integrand. We have us =ux xs + uy ys = 2x cos (x2 + y 2 )tes + 2y cos (x2 + y 2 ) =[2t2 e2s + 2(s + t)] cos [t2 e2s + (s + t)2 ] Likewise. (a) We have (b) We have 4e2(x+y) ∂ ∂ ∂ (y 2 sin xy) = sin xy ∂y (y 2 )+y 2 ∂y (sin xy) = 2y sin xy+xy 2 cos xy. t)dx a(t) b(t) =f (b(t). t) be a continuous function in the rectangle {a ≤ x ≤ b} × {c ≤ t ≤ d}. ∂y ∂ ∂ ∂2 ∂ [ex+y ]2 = ∂x e2(x+y) = 2e2(x+y) .

2 (a) Show that E (t) ≤ 0. Note that the numbers. L 1 (b) Show that E(t) ≤ 0 2 |f (x)|2 dx. Let E(t) = 1 0 u2 dx. Note that if N is any upper bound of f in D then M ≤ N. Thus. The smallest upper bound of f is called the least upper bound or the supremum. E(t) ≤ E(0) = 1 2 L L u2 (x. t)dx = k 0 u(x. t)dx − α x 0 u2 (x.4 Consider the heat problem PRELIMINARIES ut = kuxx − αu. (a) We have dE 1 = dt 2 = 0 L L 0 L ∂ 2 u (x. α > 0. If M is the supremum of f in D we write M = sup{f (x) : x ∈ D}. t) = 0 = ux (L.5 Find the supremum of f (x) = sin x. . Example 1. · · · are also upper bounds of f. t)|L − k 0 0 L u2 (x. t)dx ≤ 0 (b) From (a) we conclude that E(t) is a decreasing function of t > 0. t)ut (x. t)ux (x. k > 0. t)dx − α 0 L u2 (x.10 Example 1. 0) = L f (x). t)dx = ku(x. M + 1. 0)dx = 0 0 1 |f (x)|2 dx 2 The Least Upper Bound A function f : D → R is said to be bounded from above in D if there is a constant M such that f (x) ≤ M for all x ∈ D. t) and initial condition u(x. t)dx =−k 0 u2 (x. Solution. 0 < x < L. t)dx − α x 0 L u2 (x. t)dx ∂t L L u(x. t)uxx (x. t > 0 with boundary conditions ux (0. M + 2. We call M an upper bound of f.

1 SOME RESULTS OF CALCULUS 11 Solution. The graph of f is bounded between −1 and 1. t > 0 = 2 . sup{f (x) : x ∈ R} = 1 Example 1. t > 0 sin x sin t : x ∈ R.6 Find sup Solution. Thus. The answer is sup 2 2 sin x sin t : x ∈ R.

2π). Exercise 1.12 PRELIMINARIES Practice Problems Exercise 1. ∂t f (x)dx = 0.6 Let f be an odd function. y. t) = t ln (s ) + t3 − s4 2 3 4 (d) f (x. Exercise 1. y. y) = e3x cos y. that is.2 Find all the ﬁrst partial derivatives of the functions: √ (a) f (x. ﬁnd Exercise 1. and y = s2 t. v) = u29u +5v (f) f (x.5 In the equation ∂u ∂u − = x − 2y ∂x ∂y identify the independent variable(s) and the dependent variable. y) = x2 + ln (5x − 3y 2 ) Exercise 1. f (−x) = −f (x) for all x ∈ R.1 Compute the partial derivatives indicated: ∂ (a) ∂x (y 2 sin xy) 2 ∂2 (b) ∂x2 (ex y ) (c) ∂4 ∂x∂y 2 ∂z z ln x2 y . z) = x sin y z2 (g) f (x. x = st2 .4 If z = ex sin y. Exercise 1. Show that for all a ∈ R we have a ∂z ∂s and ∂z . Compute fx (0. z) = x2 y − 10y 2 z 3 √ 43x − 7 tan (4y) + 7 9 7 2 (c) f (s. y) = x4 + 6 y (b) f (x. −a . y) = cos x ex y−5y (e) f (u.3 Let f (x.

12 n2 t Let un (x. t) = 2 sin x sinh . .9 Let u (x. t) − 1| : x ∈ R}. f (−x) = f (x) for all x ∈ R. 0) − 1| : x ∈ R}.10 Let u (x. (b) Find sup{|un (x. Exercise 1. that is. Exercise 1. t) = 1 + e n sin nx.1 SOME RESULTS OF CALCULUS 13 Exercise 1. 2 sin x sin t . sinh x = 2 Find utt and uxx .11 Find sup 2 sinh t sin x :x∈R . Show that for all a ∈ R we have a a f (x)dx = 2 −a 0 f (x)dx. t) = u vdx. t Exercise 1. Exercise 1.7 Let f be an even function.8 Use the product rule of derivatives to derive the formula of integration by parts uv dx = uv − Exercise 1. Find utt and uxx . where ex − e−x . (a) Find sup{|un (x.

14

PRELIMINARIES

2 Sequences of Functions: Pointwise and Uniform Convergence
Later in this book we will be constructing solutions to PDEs involving inﬁnite sums of sines and cosines. These inﬁnite sums or series are called Fourier series. Fourier series are examples of series of functions. Convergence of series of functions is deﬁned in terms of convergence of a sequence of functions. In this section we study the two types of convergence of sequences of functions. Recall that a sequence of numbers {an }∞ is said to converge to a number n=1 L if and only if for every given > 0 there is a positive integer N = N ( ) such that for all n ≥ N we have|an − L| < . What is the analogue concept of convergence when the terms of the sequence are variables? Let D ⊂ R and for each n ∈ N consider a function fn : D → R. Thus, we obtain a sequence of functions {fn }∞ . For such a sequence, there n=1 are two types of convergenve that we consider in this section: pointwise convergence and uniform convergence. We say that {fn }∞ converges pointwise on D to a function f : D → R if n=1 and only if for a given a ∈ D and > 0 there is a positive integer N = N (a, ) such that if n ≥ N then |fn (a) − f (a)| < . In symbol, we write
n→∞

lim fn (a) = f (a).

It is important to note that N is a function of both a and . Example 2.1 nx Deﬁne fn : [0, ∞) → R by fn (x) = 1+n2 x2 . Show that the sequence {fn }∞ n=1 converges pointwise to the function f (x) = 0 for all x ≥ 0. Solution. For all x ≥ 0,
n→∞

lim fn (x) = lim

n→∞

nx =0 1 + n2 x2

Example 2.2 For each positive integer n let fn : (0, ∞) → ∞ be given by fn (x) = nx. Show that {fn }∞ does not converge pointwise on D. n=1

2 SEQUENCES OF FUNCTIONS: POINTWISE AND UNIFORM CONVERGENCE15 Solution. This follows from the fact that limn→∞ nx = ∞ for all x ∈ D As pointed out above, for pointwise convergence, the positive integer N depends on both the given x and . A stronger convergence concept can be deﬁned where N depends only on . Let D be a subset of R and let {fn }∞ be a sequence of functions deﬁned on n=1 D. We say that {fn }∞ converges uniformly on D to a function f : D → R n=1 if and only if for all > 0 there is a positive integer N = N ( ) such that if n ≥ N then |fn (x) − f (x)| < for all x ∈ D. This deﬁnition says that the integer N depends only on the given so that for n ≥ N , the graph of fn (x) is bounded above by the graph of f (x) + and below by the graph of f (x) − . Example 2.3 x For each positive integer n let fn : [0, 1] → R be given by fn (x) = n . Show ∞ that {fn }n=1 converges uniformly to the zero function. Solution. Let > 0 be given. Let N be a positive integer such that N > 1 . Then for n ≥ N we have 1 1 |x| ≤ ≤ < |fn (x) − f (x)| = n n N for all x ∈ [0, 1] Clearly, uniform convergence implies pointwise convergence to the same limit function. However, the converse is not true in general. Example 2.4 nx Deﬁne fn : [0, ∞) → R by fn (x) = 1+n2 x2 . By Example 2.1, this sequence converges pointwise to f (x) = 0. Let = 1 . Show that there is no positive 3 integer N with the property n ≥ N implies |fn (x) − f (x)| < for all x ≥ 0. Hence, the given sequence does not converge uniformly to f (x). Solution. For any positive integer N and for n ≥ N we have fn 1 n −f 1 n = 1 > 2

16

PRELIMINARIES

Exercise 2.1 below shows a sequence of continuous functions converging pointwise to a discontinuous function. That is, pointwise convergence does not preserve the property of continuity. One of the interesting features of uniform convergence is that it preserves continuity as shown in the next example. Example 2.5 Suppose that for each n ≥ 1 the function fn : D → R is continuous in D. Suppose that {fn }∞ converges uniformly to f. Let a ∈ D. n=1 (a) Let > 0 be given. Show that there is a positive integer N such that if n ≥ N then |fn (x) − f (x)| < 3 for all x ∈ D. (b) Show that there is a δ > 0 such that for all |x − a| < δ we have |fN (x) − fN (a)| < 3 . (c) Using (a) and (b) show that for |x − a| < δ we have |f (x) − f (a)| < . Hence, f is continuous in D since a was arbitrary. Symbolically we write
x→a n→∞

lim lim fn (x) = lim lim fn (x).
n→∞ x→a

Solution. (a) This follows from the deﬁnition of uniform convergence. (b) This follows from the fact that fN is continuous at a ∈ D. (c) For |x − a| < δ we have |f (x) − f (a)| = |f (a) − fN (a) + fN (a) − fN (x) + fN (x)−f (x)| ≤ |fN (a)−f (a)|+|fN (a)−fN (x)|+|fN (x)−f (x)| < 3 + 3 + 3 = Does pointwise convergenvce preserve integration? In real analysis, it is proven that pointwise convergence does not preserve integrability. That is, the pointwise limit of a sequence of integrable functions need not be integrable. Even when a sequence of functions converges pointwise, the process of interchanging limits and integration is not true in general. Contrary to pointwise convergence, uniform convergence preserves integration. Moreover, limits and integration can be interchanged. That is, if {fn }∞ converges uniformly to f on a closed interval [a, b] then n=1
b n→∞ b

lim

fn (x)dx =
a

a n→∞

lim fn (x)dx.

Now, what about diﬀerentiablility? Again, pointwise convergence fails in general to conserve the diﬀerentiability property. See Exercise 2.1. Does uniform convergence preserve diﬀerentiability? The answer is still no as shown in the next example.

2 SEQUENCES OF FUNCTIONS: POINTWISE AND UNIFORM CONVERGENCE17 Example 2.6 Consider the family of functions fn : [−1, 1] given by fn (x) = (a) Show that fn is diﬀerentiable for each n ≥ 1. (b) Show that for all x ∈ [−1, 1] we have
1 x2 + n .

1 |fn (x) − f (x)| ≤ √ n √ 1 1 where f (x) = |x|. Hint: Note that x2 + n + x2 ≥ √n . (c) Let > 0 be given. Show that there is a positive integer N such that for n ≥ N we have |fn (x) − f (x)| < for all x ∈ [−1, 1]. Thus, {fn }∞ converges uniformly to the non-diﬀerentiable function f (x) = n=1 |x|. Solution. (a) fn is the composition of two diﬀerentiable functions so it is diﬀerentiable with derivative −1 1 2 2 fn (x) = x x + . n (b) We have √ √ 1 1 ( x2 + n − x2 )( x2 + n + x2 ) √ 1 |fn (x) − f (x)| = x2 + − x2 = √ n x2 + 1 + x2
n

= x2 + ≤
1 n 1 √ n

1 n 1 n

+

x2

1 =√ n

1 (c) Let > 0 be given. Since limn→∞ √n = 0 we can ﬁnd a positive integer 1 N such that for all n ≥ N we have √n < . Now the answer to the question follows from this and part (b)

Even when uniform convergence occurs, the process of interchanging limits and diﬀerentiation may fail as shown in the next example.

18

PRELIMINARIES

Example 2.7 Consider the functions fn : R → R deﬁned by fn (x) = sinnnx . (a) Show that {fn }∞ converges uniformly to the function f (x) = 0. n=1 (b) Note that {fn }∞ and f are diﬀerentiable functions. Show that n=1
n→∞

lim fn (x) = f (x) = lim fn (x) .
n→∞

That is, one cannot, in general, interchange limits and derivatives. Solution. (a) Let > 0 be given. Let N be a positive integer such that N > 1 . Then for n ≥ N we have |fn (x) − f (x)| = 1 sin nx ≤ < n n

and this is true for all x ∈ R. Hence, {fn }∞ converges uniformly to the n=1 function f (x) = 0. (b) We have limn→∞ fn (π) = limn→∞ cos nπ = limn→∞ (−1)n which does not converge. However, f (π) = 0 Pointwise convergence was not enough to preserve diﬀerentiability, and neither was uniform convergence by itself. Even with uniform convergence the process of interchanging limits with derivatives is not true in general. However, if we combine pointwise convergence with uniform convergence we can indeed preserve diﬀerentiability and also switch the limit process with the process of diﬀerentiation. Theorem 2.3 Let {fn }∞ be a sequence of diﬀerentiable functions on [a, b] that converges n=1 pointwise to some function f deﬁned on [a, b]. If {fn }∞ converges uniformly n=1 on [a, b] to a function g, then the function f is diﬀerentiable with derivative equals to g. Thus, lim fn (x) = g(x) = f (x) = lim fn (x) .
n→∞

n→∞

Finally, we conclude this section with the following important result that is useful when a given sequence is bounded.

2 SEQUENCES OF FUNCTIONS: POINTWISE AND UNIFORM CONVERGENCE19 Theorem 2. We have cos x n converges uniformly to the 1 cos x | : x ∈ R} ≤ .4 Consider a sequence fn : D → R. Then this sequence converges uniformly to f : D → R if and only if n→∞ lim sup{|fn (x) − f (x)| : x ∈ D} = 0. Example 2. Solution. n n Now apply the squeeze rule for sequences we ﬁnd that 0 ≤ sup{| n→∞ lim sup{| cos x | : x ∈ R} = 0 n which implies that the given sequence converges uniformly to the zero function on R .8 Show that the sequence deﬁned by fn (x) = zero function.

Show that this sequence converges pointwise to a noncontinuous 2 function to be determined.4 Consider the sequence of functions deﬁned by fn (x) = n2 xn for all 0 ≤ x ≤ 1. Show that this sequence converges pointwise to a function f to be determined. Deﬁne f : [0. Exercise 2.2 Consider the sequence of functions fn (x) = nx + x2 n2 deﬁned for all x in R. Exercise 2. 1] → R by f (x) = 0 if 0 ≤ x < 1 1 if x = 1.20 PRELIMINARIES Practice Problems Exercise 2. n=1 (b) Show that the sequence {fn }∞ does not converge uniformly to f. Exercise 2.5) N < x < 1. . Let = 0. Hint: n=1 Suppose otherwise.5 and get a contradiction by using a point 1 (0.1 Deﬁne fn : [0. Exercise 2. 1] → R by fn (x) = xn .3 Consider the sequence of functions fn (x) = sin (nx + 3) √ n+1 deﬁned for all x in R. Show that this sequence converges pointwise to a function f to be determined. Show that this sequence does not converge pointwise to any function.5 Consider the sequence of functions deﬁned by fn (x) = (cos x)n for all − π ≤ 2 x ≤ π . (a) Show that the sequence {fn }∞ converges pointwise to f.

1).6 n Consider the sequence of functions fn (x) = x − x deﬁned on [0. 2]. Exercise 2.8 For each n ∈ N deﬁne fn : R → R by fn (x) = 1 (a) Show that fn → 2 uniformly. (b) Does fn → f uniformly on [0.9 Show that the sequence deﬁned by fn (x) = (cos x)n does not converge uniformly on [− π . 5 (b) What is the value of the limit limn→∞ 2 fn (x)dx? . ﬁnd the limit funcn=1 tion and show whether the convergence is pointwise or uniform.2 SEQUENCES OF FUNCTIONS: POINTWISE AND UNIFORM CONVERGENCE21 Exercise 2. (a) Find the pointwise limit f (x) = limn→∞ fn (x) on [0. 7 (b) Find limn→∞ 2 fn (x)dx. (b) Does {fn }∞ converge to some limit function? If so. 2].7 xn Let fn (x) = 1+xn for x ∈ [0. ﬁnd the limit funcn=1 tion and show whether the convergence is pointwise or uniform. n (a) Does {fn }∞ converge to some limit function? If so. 2n+sin2 x Exercise 2. n+cos x . 2 2 Exercise 2. π ].10 Let {fn }∞ be a sequence of functions such that n=1 sup{|fn (x)| : 2 ≤ x ≤ 5} ≤ 2n . 2]? Exercise 2. 1 + 4n (a) Show that this sequence converges uniformly to a function f to be found.

22 PRELIMINARIES .

3 The Method of Integrating Factor In this section. Indeed. we include some of the results from ODE theory that will be needed in our future discussions. y(t) = or y(t) = e− p(t)dt µ(t)g(t)dt + C C µ(t) p(t)dt 1 µ(t) e µ(t)g(t)dt + p(t)dt g(t)dt + Ce− 23 . In this chapter.1) by µ(t) and notice that the left hand side of the resulting equation is the derivative of a product. Let µ(t) = e p(t)dt . it has an antiderivative namely p(t)dt. Since p(t) is continuous.Review of Some ODEs Results Later on in this book.1) where p(t) and g(t) are continuous on the open interval a < t < b. dt Integrate both sides of the last equation with respect to t to obtain µ(t)y = Hence. we discuss a technique for solving the ﬁrst order linear nonhomogeneous equation y + p(t)y = g(t) (3. d (µ(t)y) = µ(t)g(t). Multiply Equation (3. Then techniques from the theory of ODE are required in solving the transformed ODE. we will encounter problems where a given partial diﬀerential is reduced to an ordinary diﬀerential function by means of a given change of variables.

y(1) = 5.1 Solve the initial value problem y − y = 4t. 2 The integrating factor is µ(t) = e t dt = t2 . t > 0. t Solution.2 Find the general solution to the equation 2 y + y = ln t. 0 < t < ∞ Example 3. the general solution to Equation (3. Since y(1) = 5. Example 3.1) is the sum of a particular solution of the nonhomogeneous equation and the general solution of the homogeneous equation. Multiplying the given equation by the t t integrating factor and using the product rule we notice that 1 y t = 4. Integrating with respect to t and then solving for y we ﬁnd that the general solution is given by y(t) = t 4dt + Ct = 4t2 + Ct.24 REVIEW OF SOME ODES RESULTS Notice that the second term of the previous expression is just the general solution for the homogeneous equation y + p(t)y = 0 whereas the ﬁrst term is a solution to the nonhomogeneous equation. We have p(t) = − 1 so that µ(t) = 1 . Multiplying the given equation by t2 to obtain (t2 y) = t2 ln t. t Solution. we ﬁnd C = 1 and hence the unique solution to the IVP is y(t) = 4t2 + t. That is. .

v = t3 obtaining t t2 y = Thus. dv = t2 dt. y= t t C ln t − + 2 3 9 t t3 t3 ln t − + C 3 9 . 25 The integral on the right-hand side is evaluated using integration by parts 3 with u = ln t.3 THE METHOD OF INTEGRATING FACTOR Integrating with respect to t we ﬁnd t2 y = t2 ln tdt + C. du = dt .

Exercise 3.1 Solve the IVP: y + 2ty = t.10 Solve aux + buy + cu = 0 by using the change of variables s = ax + by and t = bx − ay. Exercise 3. t Exercise 3.2 Find the general solution: y + 3y = t + e−3t .8 Suppose that y(t) = −2e−t + et + sin t is the unique solution to the IVP y + y = g(t). y(1) = − 1 exists on the 3 interval −∞ < t < ∞. . What is the value of the constant α? Exercise 3.26 REVIEW OF SOME ODES RESULTS Practice Problems Exercise 3.4 Find the general solution: y + 2y = cos (3t). Exercise 3.9 Find the value (if any) of the unique solution to the IVP y + (1 + cos t)y = 1 + cos t. Determine the functions p(t) and g(t). Exercise 3.7 Suppose that y(t) = Ce−2t + t + 1 is the general solution to the equation y + p(t)y = g(t). y(0) = 0. y(0) = y0 . Exercise 3.6 Given that the solution to the IVP ty + 4y = αt2 . t > 0.3 Find the general solution: y + 1 y = 3 cos t. y(0) = 3 in the long run? Exercise 3.5 Find the general solution: y + (cos t)y = −3 cos t. Exercise 3. Determine the constant y0 and the function g(t).

3 THE METHOD OF INTEGRATING FACTOR 27 Sample Exam Questions Exercise 3. and b is any real number.12 Show that if a and λ are positive constants.14 Solve the initial-value problem ty + 2y = t2 − t + 1.15 Solve y − 1 y = sin t. y(0) = 1 using the substitution 1 u(t) = y(t) Exercise 3. Si(t) = 0 sin s ds. y(1) = 7. Exercise 3. Exercise 3.13 Solve the initial-value problem y + y = et y 2 .11 Solve the initial value problem ty = y + t. y(1) = 1 2 Exercise 3. then every solution of the equation y + ay = be−λt has the property that y → 0 as t → ∞. Express your answer in terms of the sine t t integral. s . Hint: Consider the cases a = λ and a = λ separately. y(1) = 3.

we review the method for ordinary diﬀerentiable equations. A ﬁrst order diﬀerential equation is separable if it can be written with one variable only on the left and the other variable only on the right: f (y)y = g(t) To solve this equation. the result is generally an implicit equation involving a function of y and a function of t. In this section. substitute the values of t and y by t0 and y0 to get the value of C.2 If F is a diﬀerentiable function of y and y is a diﬀerentiable function of t and both F and y are given then the chain rule allows us to ﬁnd dF given by dt dF dF dy = · dt dy dt For separable equations.28 REVIEW OF SOME ODES RESULTS 4 The Method of Separation of Variables for ODEs The method of separation of variables that you have seen in the theory of ordinary diﬀerential equations has an analogue in the theory of partial differential equations (Section 17). Let F (t) be an antiderivative of f (t) and G(t) be an antiderivative of g(t). It may or may not be possible to solve this to get y explicitly as a function of t. Then by the Chain Rule dF dy d F (y) = = f (y)y dt dy dt Thus.” . For an initial value problem. we are given f (y)y = dF and we are asked to ﬁnd dt F (y). we proceed as follows. This process is referred to as “reversing the chain rule. f (y)y − g(t) = It follows that F (y) − G(t) = C which is equivalent to f (y)y dt = g(t)dt + C d d d F (y) − G(t) = [F (y) − G(t)] = 0 dt dt dt As you can see. Remark 4.

we ﬁnd C = 5. Separating the variables and integrating both sides we obtain y dt = y2 or − Thus.4 THE METHOD OF SEPARATION OF VARIABLES FOR ODES Example 4. − 1 = 3t2 + C y(t) d dt 1 y 6tdt dt = 6tdt 1 Since y(1) = 25 . y 2 = −4 cos (2t) + C Since y(0) = 1. The interval of existence of the solution is the interval −∞ < t < ∞ y2 2 dt = 4 sin (2t)dt . 25 Solution. we ﬁnd C = −28. y(1) = 29 1 .2 Solve the IVP yy = 4 sin (2t). This is a separable diﬀerential equation. Integrating both sides we ﬁnd d dt Thus. Now. The unique solution to the IVP is then given explicitly by 1 y(t) = 28 − 3t2 Example 4. solving explicitly for y(t) we ﬁnd √ y(t) = ± −4 cos t + 5 √ Since y(0) = 1.1 Solve the initial value problem y = 6ty 2 . y(0) = 1. Solution. we have y(t) = −4 cos t + 5.

Exercise 4.2 Solve the (separable) diﬀerential equation y = t2 y − 4y . y π yy = sin t. y( ) = −2. y(0) = 2. Exercise 4.8 Solve the IVP: 1 = 0.6 Solve the IVP: 4 sin (2t) .4 Solve the (separable) diﬀerential equation y = 2y(2 − y). 2 Exercise 4.1 Solve the (separable) diﬀerential equation y = tet 2 −ln y 2 . t+2 Exercise 4.5 Solve the IVP y = Exercise 4.7 Solve the IVP: y + Exercise 4.30 REVIEW OF SOME ODES RESULTS Practice Problems Exercise 4. . y(1) = 0. y+1 y − ty 3 = 0.3 Solve the (separable) diﬀerential equation ty = 2(y − 4). y(0) = 1. Exercise 4.

4 THE METHOD OF SEPARATION OF VARIABLES FOR ODES Exercise 4. y(0) = . y( ) = −1. 4 Exercise 4. 2 .10 Solve the IVP: 1 y = t − ty 2 .9 Solve the IVP: 31 π y = 1 + y 2 .

y0 . Exercise 4. Exercise 4. having implicit solution y 3 + t2 + sin y = 4.13 Solve the IVP (2y − sin y)y = sin t − t.32 REVIEW OF SOME ODES RESULTS Sample Exam Questions Exercise 4.11 For what values of the constants α. y(0) = 0.12 Solve the equation 3uy + uxy = 0 by using the substitution v = uy . y(0) = y0 . Exercise 4.14 State an initial value problem. Exercise 4. with initial condition imposed at t0 = 2.15 Can the diﬀerential equation dy = x2 − xy dx be solved by the method of separation of variables? Explain. and integer n is the function y(t) = 1 (4 + t)− 2 a solution of the initial value problem? y + αy n = 0. .

Example 5. Hence. Thus. or two conjugate complex solutions depending on the sign of the discriminant b2 − 4ac. and Laplace’s equations using the method of separation of variables for PDEs one ends up confronting second order linear ODEs. One such function is y(t) = ert . The characteristic equation is a quadratic equation. y(0) = 1. We call (5. The general solution to (5. Thus.2) the characteristic equation for (5. it is deemed necessary to review some of the techniques usided in solving second order linear ordinary diﬀerential equations which we do in this section. Notice ﬁrst that for b = 0 and c = 0 the function y is a constant multiple of y. We start ﬁrst by considering the second order linear ODE with constant coeﬃcients given by ay + by + cy = 0 (5. So it makes sense to look for a function with such property. a function y(t) = ert is a solution to (5.1) leads to ay + by + cy = ar2 ert + brert + cert = (ar2 + br + c)ert = 0 Since ert > 0 for all t. Substituting this function into (5.5 SECOND ORDER LINEAR ODES 33 5 Second Order Linear ODEs When solving second order partial diﬀerential equations such as the heat. the previous equation leads to ar2 + br + c = 0 (5. √ b2 In this case.1) where a. . we consider the following three cases: Case 1: b2 − 4ac > 0. two equal solutions. y (0) = 2. b and c are constants with a = 0.2).1) when r satisﬁes equation (5. equation (5. wave.1) and the polynomial C(r) = ar2 + br + c is called the characteristic polynomial.2) Thus. this equation can have two distinct real solutions.1 Solve the initial value problem y − y − 6y = 0.2) have two distinct real roots r1 = −b− 4a −4ac and √ b2 r2 = −b+ 4a −4ac .1) is given by y(t) = c1 er1 t + c2 er2 t where c1 and c2 are arbitrary constants. Describe the behavior of the solution y(t) as t → −∞ and t → ∞.

The characteristic polynomial is C(r) = r2 − r − 6 = (r − 3)(r + 2) so that the characteristic equation r2 − r − 6 = 0 has the solutions r1 = 3 and r2 = −2. y(t) → ∞. 5 . the general solution is given by y(t) = c1 e−t + c2 te−t . Thus.2 Solve the initial value problem: y + 2y + y = 0. the characteristic equation has the single root r = − 2a . e3t → 0 and e−2t → ∞. Similarly. The general solution is then given by y(t) = c1 e3t + c2 e−2t . the unique solution to the initial value problem is 1 y(t) = (4e3t + e−2t ). 5 As t → −∞. The general solution to (5. Example 5. Solving this system by the method of elimination we ﬁnd c1 = Hence. The characteristic equation r2 +2r+1 = 0 has a repeated root: r1 = r2 = −1. y1 (0) = −1. b b 4 5 and c2 = 1 . b In this case. y(0) = 1. The conditions y(0) = 1 and y (0) = 2 lead to the system c1 + c2 = 1 3c1 − 2c2 = 2. Taking the derivative to obtain y (t) = 3c1 e3t − 2c2 e−2t . y(t) → ∞ as t → ∞ Case 2: b2 − 4ac = 0. Solution. Thus.1) is given by y(t) = c1 e− 2a t + c2 te− 2a t where c1 and c2 are arbitrary constants.34 REVIEW OF SOME ODES RESULTS Solution.

2 = 5 ± 2i. y (0) = 3. The general solution is given by y(t) = eαt (c1 cos βt + c2 sin βt) b where α = − 2a . The characteristic equation r2 − 10r + 29 = 0 has the complex roots r1. the complex roots of equation (5. Hence. The initial conditions yield c1 = 1 and c2 = −1. Thus.1) are given by √ −b ± i 4ac − b2 r1. the general solution is given by the expression y(t) = e5t (c1 cos 2t + c2 sin 2t). In this case. a < x < b. β = √ 4ac−b2 .5 SECOND ORDER LINEAR ODES 35 The two conditions y(0) = 1 and y (0) = −1 lead to c2 = 1 and c1 = 0. Example 5. y(0) = 1. the unique solution is y(t) = e−t Case 3: b2 − 4ac < 0. 2a and c1 and c2 are real numbers. Finding y we obtain y (t) = e5t [(5c1 + 2c2 ) cos 2t + (5c2 − 2c1 ) sin 2t].3 Solve the initial value problem y − 10y + 29y = 0. dx2 . Solution. Thus. the unique solution to the initial value problem is y(t) = e5t (cos 2t − sin 2t) The Eigenvalue Problem Consider the question of ﬁnding a twice diﬀerentaible function u satisfying the ordinary diﬀerential equation d2 u = λu.2 = 2a √ where i = −1.

This problem is referred to as the eigenvalue problem for the following reason: Deﬁne the function d2 L ≡ dx2 . c are constants is called an Euler equation. Again. suppose that λ < 0. u ≡ 0. To solve Euler equation. Then u(x) = Ae conditions imply that u ≡ 0. Finally.36 REVIEW OF SOME ODES RESULTS subject to the boundary conditions u(a) = u(b) = 0. Then the given equation can be written as Lu = λu. Using the condition u(0) = 0 to obtain A = 0. Then u(x) = ax+b for arbitrary constants a and b. In linear algebra. the boundary Suppose that λ > 0. Using the boundary conditions we ﬁnd a √ b = 0. Hence. Suppose ﬁrst that λ = 0. λ is called an eigenvalue of L with corresponding eigenvector u. Then u(x) = A cos −λx + B sin −λx. √ √ Now. and complex conjugate roots. using the principle of superposition we ﬁnd that the general solution to the eigenvalue problem is given by ∞ u(x) = n=1 An un (x) where the convergence is pointwise convergence (See Section 2). We call λn an eigenvalue with corresponding eigenfunction un (x) = An sin nπx. b. Diﬀerent solutions to the eigenvalue problem are obtained depending on the sign of λ. This happens when λ = λn = −(nπ)2 where n ∈ N. one starts with solutions of the form y = xr (with x > 0) where r is to be determined. Plugging this into the diﬀerential equation to get ax2 r(r − 1)xr−2 + bxrxr−1 + cxr =0 (ar2 − ar + br + c)xr =0 ar2 − (a − b)r + c =0 This last equation is a quadratic equation in r and so we will have three cases to look at : Real distinct roots. Using the condition u(1) = 0 and assuming we are looking for non-trivial solution u we expect to have √ sin −λ = 0. Euler Equations A second order linear diﬀerential equations of the form ax2 y + bxy + cy = 0 where a. . = √ λx − λx + Be . double roots.

Example 5. Hence. If the quadratic equation has two complex conjugate solutions r1. the unique solution is given by y= 2 5 2 x 2 − x−3 11 11 5 Second Order Linear Inhomogeneous ODE: The Method of Undetermined Coeﬃcients We consider the nonhomogeneous second order ay + by + cy = g(t). Solving this system of two unknowns we ﬁnd A = 11 and 2 2 B = − 11 . Solution. a < t < b.4 Solve the initial value problem 2x2 y + 3xy − 15y = 0 y(1) = 0. The condition y (1) = 1 implies 5 2 A − 3B = 1. The condition y(1) = 0 implies A + B = 0. Hence. If the quadratic equation has two equal roots r1 = r2 = r then the general solution is given by y(x) = xr (A + B ln x). Letting y = xr we obtain the quadratic equation 2r2 + r − 15 = 0 whose 5 roots are r1 = 2 and r2 = −3. We know that the general solution has the structure y(t) = c1 y1 (t) + c2 y2 (t) + yp (t) .2 = α ± iβ then the general solution is given by y(x) = xα (A cos (β ln x) + B sin (β ln x)).5 SECOND ORDER LINEAR ODES 37 If the quadratic equation has two distinct real roots r1 and r2 then the general solution is given by y(x) = Axr1 + Bxr2 . the general solution is given by y(x) = Ax 2 + Bx−3 . y (1) = 1.

or 2. and polynomials an tn + an−1 tn−1 + · · · a1 t + a0 . Form of g(t) an tn + an−1 tn−1 + · · · + a1 t + a0 [an tn + an−1 tn−1 + · · · + a1 t + a0 ]eαt [an tn + an−1 tn−1 + · · · + a1 t + a0 ] cos αt or [an tn + an−1 tn−1 + · · · + a1 t + a0 ] sin αt eαt [an tn + an−1 tn−1 + · · · + a1 t + a0 ] sin βt or eαt [an tn + an−1 tn−1 + · · · + a1 t + a0 ] cos βt Form of yp (t) tr [An tn + An−1 tn−1 + · · · + A1 t + A0 tr [An tn + An−1 tn−1 + · · · + A1 t + A0 ]eαt tr [(An tn + An−1 tn−1 + · · · + A1 t + A0 ) cos αt +(Bn tn + Bn−1 tn−1 + · · · + B1 t + B0 ) sin αt] tr [(An tn + An−1 tn−1 + · · · + A1 t + A0 )eαt cos βt +(Bn tn + Bn−1 tn−1 + · · · + B1 t + B0 )eαt sin βt] The number r is chosen to be the smallest nonnegative integer such that no term in the assumed form is a solution of the homogeneous equation ay + by + cy = 0. This is possible only for special functions g(t). (Note that if both cosine and sine terms are present. (b) y + 4y = te2t cos t. One way to ﬁnding yp is by using the methos of undetermined coeﬃcients. The value of r will be 0. But if they have diﬀerent arguments they must be treated separately. (d) y + 4y = t2 cos 2t. . In the following table we list examples of g(t) along with the corresponding form of the particular solution. cos (kt). 1.38 REVIEW OF SOME ODES RESULTS where yp (t) is a particular solution to the nonhomogeneous equation. We will assume that g(t) being simple means it is some combination of terms like ert . each resulting in a combination of sine and cosine terms in yp . but these special cases arise quite frequently in applications. We will write y(t) = yh (t) + yp (t) where yh (t) = c1 y1 (t) + c2 y2 (t). (c) y + 4y = 2t2 + 5 sin 2t + e3t .) Based on those terms we will put together a candidate yp that has some constants in it we need to solve for: Those are the undetermined coeﬃcients this method is named for. The idea behind the method of undetermined coeﬃcients is to look for yp (t) which is of a form like that of g(t). Example 5. sin (kt).5 List an appropriate form for a particular solution of (a) y + 4y = t2 e3t . if they have the same argument kt they can be treated as one.

6 Find the general solution of y − 2y − 3y = 4t − 5 + 6te2t Solution. (a) For g(t) = t2 e3t . From 2C − 3D = 0 we ﬁnd 4 D = − 3 . Thus. −2A−3B = −5 23 so that B = 9 . Thus yp (t) = (A2 t2 + A1 t + A0 )e3t (b) An appropriate form is yp (t) = tr [(A1 t + A0 )e2t cos t + (B1 t + B0 )e2t sin t] We take r = 0 since no term in the assumed form for yp is present in the expression of yh (t). Also. Since −3C = 6 we ﬁnd C = −2. Thus yp (t) = (A1 t + A0 )e2t cos t + (B1 t + B0 )e2t sin t (c) yp (t) = A2 t2 + A1 t + A0 + B0 t cos 2t + C0 t sin 2t + D0 e3t (d) yp (t) = t(A2 t2 + A1 t + A0 ) cos 2t + t(B2 t2 + B1 t + B0 ) cos 2t Example 5.5 SECOND ORDER LINEAR ODES 39 Solution. The general solution to the homogeneous equation is yh (t) = c1 cos t+c2 sin t. It follows that 4 23 4 y(t) = c1 e−t + c2 te3t − t + − 2t + 3 9 3 e2t . an appropriate particular solution has the form yp (t) = tr (A2 t2 + A1 t + A0 )e3t . Inserting this into the diﬀerential equation leads to −3At − 2A − 3B − 3Cte2t + (2C − D)e2t = 4t − 5 + 6te2t 4 From this identity we obtain −3A = 4 so that A = − 3 . We take r = 0 since no term in the assumed form for yp is present in the expression of yh (t). yh (t) = c1 e−t + c2 te3t A guess for the particular solution is yp (t) = At + B + Cte2t + De2t . The characteristic equation of the homogeneous equation is r2 − 2r − 3 = 0 with roots r1 = −1 and r2 = 3.

e− 2 ) and (2. y(0) = −2.40 REVIEW OF SOME ODES RESULTS Practice Problems Exercise 5. y(3) = −2. Exercise 5. Exercise 5. y (0) = 1 Describe the behavior of the solution y(t) as t → −∞ and t → ∞. y (3) = − Exercise 5. 5 3 .3 Solve the initial value problem 2y − y = 0. Determine y(0) and y (0).2 Solve the initial value exercise y + 4y + 2y = 0.4 Find a homogeneous second-order linear ordinary diﬀerential equation whose general solution is y(t) = c1 e2t + c2 e−t . Exercise 5. 0). y(0) = −1. y(0) = 0.5 Solve the IVP 9y − 6y + y = 0. y (0) = 4 Describe the behavior of the solution y(t) as t → −∞ and t → ∞. y(5) = 4e−2 . Exercise 5. y (5) = − e−2 5 Exercise 5.7 The graph of a solution y(t) of the diﬀerential equation 4y + 4y + y = 0 1 passes through the points (1. y (0) = √ 2 Describe the behavior of the solution y(t) as t → −∞ and t → ∞.1 Solve the initial value problem y − 4y + 3y = 0.6 Solve the IVP 3 25y + 20y + 4y = 0.

Exercise 5.12 Find the general solution of y + 4y − 2y = 2t2 − 3t + 6 41 .9 Solve the IVP y + 2y + 2y = 0. y (0) = −1 Exercise 5.5 SECOND ORDER LINEAR ODES Exercise 5.10 Solve the IVP 2y − 2y + y = 0. y (−π) = −1 Exercise 5.11 Find the general solution of y − y + y = 2 sin 3t Exercise 5. y(−π) = 1.8 Find the general solution of y − 6y + 9y = 0. y(0) = 3.

x2 y − 7xy + 16y = 0.16 Show by direct computation that the eigenvalue problems (ky (x)) + λy(x) = 0 with the following boundary conditions have no negative eigenvalues λ : (a) y(0) = y(L) = 0 (b) y (0) = y (L) = 0 (c) y(L) = y(−L).42 REVIEW OF SOME ODES RESULTS Sample Exam Questions Exercise 5.19 Solve using undetermined coeﬃcients: y + y − 2y = t + sin 2t. Exercise 5.14 Find the general solution to the following diﬀerential equation. y (L) = y (−L) and k(−L) = k(L).17 Solve the initial-value problem: 2y + 5y − 3y = 0. y (0) = 0 . Exercise 5. x2 y + 3xy + 4y = 0.13 Find the general solution to the following diﬀerential equation. y (0) = 1.15 Consider the diﬀerential equation d2 y + λy = 0. y(0) = 2.18 Find the general solution of y − y = 5et − sin 2t Exercise 5. Exercise 5. dx2 Determine the eigenvalues λ and the corresponding eigenfunctions if y satisﬁes the following boundary conditions: (a) y(0) = y(π) = 0 (b) y(0) = y (L) = 0 (c) y (0) = y(1) = 0. Exercise 5. Exercise 5. y(0) = 1.

43 .2) ∂t ∂x ∂y If the unknown function is a function in one single variable then the diﬀerential equation is called an ordinary diﬀerential equation. The purpose of this chapter is to introduce the reader to the basic terms of partial diﬀerential equations. in short PDE. So the study of partial diﬀerential equation is of great importance to the above mentioned ﬁelds. dy d2 y − 5 + 3y = −3 (6.1). In contrast. In this book we will be focusing on partial diﬀerential equations.Introduction to PDEs Many ﬁelds in engineering and the physical sciences require the study of ODE and PDE. aerodynamics. oceanography. A diﬀerential equation is an equation that involves an unknown scalar function (the dependent variable) and one or more of its derivatives. geophysics (seismic wave propagation). when the unknown function is a function of two or more independent variables then the diﬀerential equation is called a partial diﬀerential equation.1) 2 dx dx or ∂u ∂ 2 u ∂ 2 u − 2 − 2 + u = 0. optics. Examples of those ﬁelds are acoustics. An example of an ordinary diﬀerential equation is Equation (6. plasma physics (ionized liquids and gases). Equation (6. (6. 6 The Basic Concepts The goal of this section is to introduce the reader to the basic concepts and notations that will be used in the remainder of this book. meteorology. electrodynamics.2) is an example of a partial diﬀerential equation. ﬂuid dynamics. For example. heat transfer. petroleum engineering. elasticity. quantum mechanics.

dt ∂y (d) ∂u − 4 ∂y = u + 3y.44 INTRODUCTION TO PDES Example 6. (a) Independent variable is x and the dependent variable is y. ∂v Solution.2) is a second order partial diﬀerential equation.1 Identify which variables are dependent variable or independent variable(s) for the following diﬀerential equations. (a) PDE (b) ODE (c) PDE The order of a partial diﬀerential equation is the highest order derivative occurring in the equation. Solution. d4 (a) dxy − x2 + y = 0 4 (b) utt + xutx = 0. (d) Independent variable are u and v and the dependent variable is y Example 6. (c) ut + cux = 5. t) (d) ut + uux + uxxx = 0 (e) utt + uxxxx = 0. Example 6. (b) Independent variables are x and t and the dependent variable is u. Thus.2 Classify the following as either ODE or PDE.3 Find the order of each of the following partial diﬀerential equations: (a) xux + yuy = x2 + y 2 (b) uux + uy = 2 (c) utt − c2 uxx = f (x. (a) First order (b) First order (c) Second order (d) Third order (e) Fourth order . (a) ut = c2 uxx . (b) y − 4y + 5y = 0. (c) Independent variable is t and the dependent variable is x. (c) x dx = 4. Solution. (6.

For example. y)uxy + C(x. y)ux + E(x. y)uxx + B(x. y. u2 + 2uxy = 0. y. uy )uxy +C(x. y) whereas a second order linear partial diﬀerential equation has the form A(x. uy ). For example. y. .6 THE BASIC CONCEPTS 45 A partial diﬀerential equation is called linear if it is linear in the unknown function and all its derivatives with coeﬃcients depend only on the independent variables. u)ux + B(x. A partial diﬀerential equation that is not linear is called nonlinear. y. a ﬁrst order quasi-linear partial diﬀerential equation has the form A(x. ux . y)uy = C(x. y)uxx + B(x. y)ux + B(x. For example. u) whereas a second order semi-linear partial diﬀerential equation has the form A(x. ux . linear PDEs are usually simpler to analyze/solve than nonlinear PDEs. y)uy + F (x. u)uy = C(x. u. y)u = D(x. y. u. y)u = G(x. a ﬁrst order semi-linear partial diﬀerential equation has the form A(x. ux . y). u. ux . uy )uxx +B(x. y)uy + C(x. y)uxy + C(x. A partial diﬀerential equation is semi-linear if it is quasi-linear and the coeﬃcients of the highest-order derivatives are functions of independent variables only. y. y. A partial diﬀerential equation is called quasi-linear if it is linear in the highest-order derivatives which appear in the equation (regardless of the manner in which lower-order derivatives and unknown functions occur in the equation). u. y. y)uyy + D(x. Note that linear and semi-linear partial diﬀerential equations are special cases of quasi-linear equations. a ﬁrst order linear partial diﬀerential equation has the form A(x. u. y)uyy = D(x. For example. y. uy )uyy = D(x. uy ). ux . x As for ODEs. y)ux + B(x. u) whereas a second order quasi-linear partial diﬀerential equation has the form A(x.

semilinear. semilinear. Linearity imposes two key requirements: L[u + v] = L[u] + L[v] and L[αu] = αL[u]. tt Solution. semilinear. nonlinear.46 INTRODUCTION TO PDES Example 6. (a) Linear. v and any constant α. t) (d) ut + uux + uxxx = 0 (e) u2 + uxxxx = 0. The operator L is assembled by summing the basic partial derivative operators. with coeﬃcients depending on the independent variables. (c) Linear. quasilinear. there is a term in the equation that involves only . nonlinear. or nonlinear: (a) xux + yuy = x2 + y 2 (b) uux + uy = 2 (c) utt − c2 uxx = f (x.5 Deﬁne a linear diﬀerential operator for the PDE ut = c2 uxx . (e) Quasilinear. nonlinear A more precise deﬁnition of a linear diﬀerential equation begins with the concept of a linear diﬀerential operator L. Solution. (b) Quasilinear. A linear partial diﬀerential equation that is not homogeneous is called nonhomogeneous. The operator acts on suﬃciently smooth functions depending on the relevant independent variables. (d) Quasilinear. Let L[u] = ut − c2 uxx . for any two (suﬃciently smooth) functions u. Example 6. quasilinear. In this case. semilinear. semilinear.4 Determine whether the given PDE is linear. Then one can easily check that L[u + v] = L[u] + L[v] and L[αu] = αL[u] A linear partial diﬀerential equation is called homogeneous if every term of the equation involves the unknown function or its partial derivatives. quasilinear.

Polar coordinates r. Both time and space coordinates are independent variables. processes. while x. y. . (c) Homogeneous 47 Finally. θ will also be used when needed.6 THE BASIC CONCEPTS the independent variables. (b) Homogeneous. we shall be employing a few basic notational conventions regarding the variables that appear in our diﬀerential equations. meaning time-varying.6 Determine whether the equation is homogeneous or nonhomogeneous: (a) xux + yuy = x2 + y 2 . and our notational conventions appear at the appropriate places in the exposition. The time variable t appears when modeling dynamical . and so only involves the space variables. An equilibrium equation models an unchanging physical system. A homogeneous linear partial diﬀerential equation has the form L[u] = 0 where L is a linear diﬀerential operator. (b) utt = c2 uxx . (a) Nonhomogeneous because of x2 + y 2 . Example 6. Solution. z will represent (Cartesian) space coordinates. (c) uxx + uyy = 0. We always use t to denote time.

2 Write the equation uxx + 2uxy + uyy = 0 in the coordinates s = x.1 Classify the following equations as either ODE or PDE. Exercise 6.3 Write the equation uxx − 2uxy + 5uyy = 0 in the coordinates s = x + y. .5 For each of the following PDEs. (a) x2 uxx + ex u = xuxyy (b) ey uxxx + ex u = − sin y + 10xuy (c) y 2 uxx + ex uux = 2xuy + u (d) ux uxxy + ex uuy = 5x2 ux (e) ut = k 2 (uxx + uyy ) + f (x.4 For each of the following PDEs. (a) (y )4 + (b) ∂u ∂x t2 (y )2 +4 =0 + y ∂u = ∂y y−x y+x (c) y − 4y = 0 Exercise 6. t). For linear PDEs. determine its order and whether it is linear or not. also state whether it is homogeneous or inhomogeneous: (a) uux + x2 uyyy + sin x = 0 2 (b) ux + ex uy = 0 (c) utt + (sin y)uyy − et cos y = 0. For nonlinear PDEs. Exercise 6. y. state its order and whether it is linear or nonlinear. If it is linear. circle all term(s) that are not linear. t = 2x. Exercise 6. t = x − y.48 INTRODUCTION TO PDES Practice Problems Exercise 6. state also whether the equation is homogeneous or not.

2 2 (c) Minimal surface equation: (1 + Zy )Zxx − 2Zx Zy Zxy + (1 + Zx )Zyy = 0. Exercise 6.8 Write down the general form of a linear ﬁrst order diﬀerential equation of a function in three variables. (c) The thin ﬁlm equation for h(x.6 THE BASIC CONCEPTS 49 Exercise 6. . where ρ(x. (e) The Poisson Equation for the electric potential Φ(x. y. z) is a known charge density. z). t) : ut = kuxx . y. y. (b) Convection (transport) equation: ut + cux = 0. linear or nonlinear. z) : Φxx + Φyy + Φzz = 4πρ(x. t) : ht = −(hhxxx )x .7 Classify the following diﬀerential equations as ODEs or PDEs. t) : wtt = c2 wxx . (a) The diﬀusion equation for u(x. (b) The wave equation for w(x. Exercise 6.6 Which of the following PDEs are linear? (a) Laplace’s equation: uxx + uyy = 0. (d) Korteweg-Vries equation: ut + 6uux = uxxx . (d) The forced harmonic oscillator for y(t) : ytt + ω 2 y = F cos (ωt). determine whether or not they are homogeneous. (f) Burger’s equation for h(x. For the linear equations. t) : ht + hhx = νhxx . and determine their order.

and classify them as linear or nonlinear. . y) = x to show that uww = 0.10 Consider the second-order PDE uxx + 4uxy + 4uyy = 0. (a) x2 uxxy + y 2 uyy − log (1 + y 2 )u = 0 (b) ux + u3 = 1 (c) uxxyy + ex ux = y (d) uuxx + uyy − u = 0 (e) uxx + ut = 3u.9 Give the orders of the following PDEs.50 INTRODUCTION TO PDES Exercise 6. Exercise 6. specify whether it is homogeneous or inhomogeneous. If the PDE is linear. y) = y − 2x and w(x. Use the change of variables v(x.

a. y) = x + y. Assume a2 +b2 > 0.6 THE BASIC CONCEPTS 51 Sample Exam Questions Exercise 6. Exercise 6. Exercise 6. Exercise 6.11 Write the one dimensional wave equation utt = c2 uxx in the coordinates v = x + ct and w = x − ct.15 Write the PDE aut + bux = u. . y) = ax+by and t(x. y) = bx−ay. y) = y − 3x and w(x.14 Write the PDE ux + uy = 1 in the coordinates s = x + y and t = x − y. b = 0 1 in the coordinates v = ax − bt and w = a t.13 Write the PDE aux + buy = 0 in the coordinates s(x. Exercise 6.12 Write the PDE uxx + 2uxy − 3uyy = 0 in the coordinates v(x.

To solve a PDE is to ﬁnd all its classical solutions. Example 7. A formula that expresses all the solutions of a PDE is called the general solution of the equation. y. Solution. Consider the equation x2 − y 2 = 0. y) = f (x) + g(y). there are solutions that do not require the smoothness property. the general solution of a partial diﬀerential equation is an expression that involves arbitrary functions. In the case of only two independent variables x and y. u) space. This is in contrast to the general solution of an ordinary diﬀerential equation which involves arbitrary constants. in the theory of nonlinear pdes. t) = e−λ α t (cos λx − sin λx) is a solution to the equation ut − α2 uxx = 0. a classical solution enjoys properties such as smootheness (i. Solution. a solution u(x. In general. where g is an arbitrary diﬀerentiable function Note that the general solution in the previous example involves two arbitrary functions.1 2 2 Show that u(x.52 INTRODUCTION TO PDES 7 Solutions and Related Topics By a classical solution or strong solution to a partial diﬀerential equation we mean a function that satisﬁes the equation. Usually. Integrating ux with respect to x we ﬁnd u(x. called a solution surface or an integral surface in the (x.2 Find the general solution of uxy = 0. Integrating ﬁrst we respect to y we ﬁnd ux (x. . We illustrate this concept using equations rather than pdes. y) = f (x). Since ut − α2 uxx = −λ2 α2 e−λ 2 α2 t (cos λx − sin λx) − α2 (−λ2 cos λx + λ2 sin λx) = 0 the given function is a solution to the given equation Example 7. differentiability) and continuity.e. y) is visualized geometrically as a surface. Such solutions are called weak solutions or generalized solutions. where f is an arbitrary diﬀerentiable function. However.

Proof.1 The sum of two solutions to a homogeneous linear diﬀerential equation is again a solution. we conclude that u is a classical solution to the given PDE Now. as is the product of a solution by any constant. L[u1 + u2 ] = L[u1 ] + L[u2 ] = 0. un are solutions to a common homogeneous linear partial diﬀerential equation L[u] = 0.e. uxx exist and are continuous in D(i. u is smooth in D) and u satisﬁes equation (7. This solution is inﬁnitely diﬀerentiable function. Let u1 . In this book. · · · . and so the scalar multiple αu is also a solution The following result is known as the superposition principle for homogeneous linear equations. ux .2 If u1 . the word solution will refer to a classical solution.7 SOLUTIONS AND RELATED TOPICS 53 The function y = x is a classical solution of this equation. and u any solution. Assume that the domain of deﬁnition of u is D ⊂ R2 . consider the linear diﬀerential operator L as deﬁned in the previous section. if α is any constant. utx . meaning that L[u1 ] = 0 and L[u2 ] = 0. u2 be solutions. and hence their sum u1 + u2 is a solution. thanks to linearity. The deﬁning properties of linearity immediately imply the key facts concerning homogeneous linear (diﬀerential) equations.1) Solution. this solution is not diﬀerentiable at 0. ut . Since u. Theorem 7. t) = t + 1 x2 is a classical solution to the PDE 2 ut = uxx . Theorem 7. Then. We call such a solution a weak solution. then the linear combination u = c1 u1 + · · · + cn un is a solution for any choice of constants c1 .. then L[αu] = αL[u] = α0 = 0.3 Show that u(x. Similarly. the function y = |x| is also a solution to the given equation. . · · · . Example 7. cn . (7. On the other hand.1). However.

54 INTRODUCTION TO PDES Proof. Proof. L[uh ] = L[u − ui ] = L[u] − L[ui ] = 0. by linearity. where uh is an arbitrary solution to the corresponding homogeneous equation L[uh ] = 0. To show that every solution to the inhomogeneous equation can be expressed in this manner. Step two is to ﬁnd a particular solution to the inhomogeneous version. · · · . if the functions are solutions. L[u] =L[c1 u1 + · · · + cn un ] = L[c1 u1 + · · · + cn−1 un−1 ] + L[cn un ] = · · · = L[c1 u1 ] + · · · + L[cn un ] = c1 L[u1 ] + · · · + cn L[un ]. proving that u is also a solution to the homogeneous equation L[u] = 0 In physical applications. cn . You already learned the basic philosophy for solving of inhomogeneous linear equations in your study of elementary ordinary diﬀerential equations. External forcing is represented by an additional term that does not involve the dependent variable. In particular. and f is a given non-zero function of the independent variables alone. then the right hand side of the above equation vanishes. This results in the nonhomogeneous equation L[u] = f where L is a linear partial diﬀerential operator. L[u] = L[ui + uh ] = L[ui ] + L[uh ] = f + 0 = f. Let us ﬁrst show that u = ui + uh is also a solution whenever L[u] = 0. By linearity. Then. un and any constants c1 . u is the dependent variable. Set uh = u − ui . · · · . The key fact is that. · · · .3 Let vi be a particular solution to the inhomogeneous linear equation L[ui ] = f. L[un ] = 0. homogeneous linear equations model unforced systems that are subject to their own internal constraints. Here is the general version of this procedure: Theorem 7. . for any suﬃciently smooth functions u1 . thanks to the linearity of L. Step one is to determine the general solution to the homogeneous equation. The general solution to the inhomogeneous equation is then obtained by adding the two together. suppose u satisﬁes L[u] = f. so L[u1 ] = 0. Then the general solution to L[u] = f is given by u = ui + uh .

7 SOLUTIONS AND RELATED TOPICS 55 and hence uh is a solution to the homogeneous diﬀerential equation. the dependent function u is prescribed on the boundary of the bounded domain. a general solution of a partial diﬀerential equation has inﬁnitely many solutions. For example. For example. even ﬁnding one solution to a nonlinear partial diﬀerential equation can be quite a challenge. As you have noticed by now. y) =0 if 0 < x. the . to appreciate some recent developments in this fascinating area of contemporary research and applications. uxx + uyy =0 u(x. y < 1 if 0 < x < 1 if 0 < y < 1. These conditions determine the unique solution of interest. knowledge of several solutions is of scant help constructing others. one can interpret the particular solution ui as a response of the system to the external forcing function. But we will have occasion to brieﬂy foray into the nonlinear realm. 1) =0 ux (0. for example. usually called initial or boundary conditions. nonlinear equations are much tougher to deal with. A boundary value problem is a partial diﬀerential equation where either the unknown function or its derivatives have values assigned on the physical boundary of the domain in which the problem is speciﬁed. As observed above. In almost all cases. Indeed. u = ui + uh has the required form In physical applications. There are three types of boundary conditions which arise frequently in formulating physical problems: 1. Dirichlet Boundary Conditions: In this case. In this introductory course. These conditions are called boundary conditions. y) = ux (1. if the bounded domain is the rectangular plate 0 < x < L1 and 0 < y < L2 . one solution of a linear PDE leads to the creation of lots of solutions. this general solution is of little use since it has to satisfy other supplementary conditions. Thus. while the solution uh to the homogeneous equation represents the system’s internal. The general solution to a linear inhomogeneous equation is thus a combination of the external and internal responses. we will primarily − but not exclusively − concentrate on analyzing the most basic linear partial diﬀerential equations. In contrast. 0) = u(x. unforced motion.

where α and β are constants. y). 0) =0 Clearly. t) = α and ux (L. and u(x. t) = . namely the continuous dependence is important in physical problems. The boundary conditions are called homogeneous if the dependent variable is zero at any point on the boundary. Robin or mixed Boundary Conditions: This occurs when the dependent variable and its ﬁrst partial derivatives are prescribed on the boundary of the bounded domain. We say that an initial and/or boundary value problem associated with a PDE is well-posed if it has a solution which is unique and depends continuously on the data given in the problem. Example 7. An initial value problem (or Cauchy problem) is a partial diﬀerential equation together with a set of additional conditions on the unknwon function or its derivatives at a point in the given domain of the solution. Show that the function u (x. ﬁrst partial derivatives are prescribed on the boundary of the bounded domain. 0) =f (x). (a) Let 0 < << 1 be a very small number. the Neuman boundary conditions for a rod of length L. 0) = ut (x. For example. Neumann Boundary Conditions: In this case. otherwise the boundary conditions are called nonhomogeneous. For example. t) = β. where 0 < x < L. Such solutions are said to be stable. 2. t) =0 u(x. the transport equation ut (x. t) + cux (x. L2 ) are prescribed. are of the form ux (0.4 For x ∈ R and t > 0 we consider the initial value problem utt − uxx =0 u(x. The last condition. u(x. It can be shown that initial conditions for a pde are necessary and suﬃcient for the existence of a unique solution. These conditions are called initial value conditions. u(x. t) = 0 is a solution to this problem. 3. u(L1 .56 INTRODUCTION TO PDES boundary conditions u(0. 0). y). This condition means that the solution changes by a small amount when the conditions change a little.

Thus. sup{|u (x. t)| : x ∈ R. 0) =0 . 0) = sin t x . t)| : x ∈ R. sup{|u (x. 0) = ut (x. the initial value problem is well-posed. u (x. Solution. t) − u(x. 0) = sin x (b) Show that for a ﬁxed t > 0. ∂∂tu − 2 (b) We have 2 = sin = − sin = cos = − sin x x x x cos sin sin sin t t t t ∂2u ∂x2 = 0. t > 0} A problem that is not well-posed is referred to as an ill-posed problem. Hence.5 For x ∈ R and t > 0 we consider the initial value problem utt + uxx =0 u(x. Moreover. t > 0} = 2 sup{ sin = 2 x sin : x ∈ R.7 SOLUTIONS AND RELATED TOPICS 2 57 sin x sin t is a solution to the problem utt − uxx =0 u(x. 0) =0 ut (x. Example 7. We illustrate this concept in the next example. 0) = 0 and ∂ u ∂t (x. t) − u(x. t > 0} = 2 . a small change in the initial data leads to a small change in the solution. (a) We have ∂u ∂t 2 ∂ u ∂t2 ∂u ∂x ∂ 2u ∂x2 Thus.

t) − u(x. t) − ∂ (b) Show that sup{| ∂t u (x. 0)| : x ∈ R} = sup{ sin ∂t x = sup{ sin t : x ∈ R} = x and sup{|u (x. ∂ x u (x. Solution. u(x. t) = 2 sin x sinh t . 0) = 0 and (x. (a) We have ∂u ∂t 2 ∂ u ∂t2 ∂u ∂x ∂ 2u ∂x2 Thus. ∂∂tu + 2 (b) We have sup{| 2 = sin = sin = cos = − sin x x x x cosh sinh sinh sinh t t t t ∂ u ∂t ∂2u ∂x2 = 0. 0) = sin : x ∈ R} x .58 INTRODUCTION TO PDES Clearly. Moreover. 0) =0 ut (x. 0)| : x ∈ R} = 2 sinh t . t)| : x ∈ R} = (c) Find limt→∞ sup{|u (x. u (x. t) − u(x. 0) = sin x and sup{|u (x. (a) Let 0 < << 1 be a very small number. 0) − ut (x. where ex − e−x sinh x = 2 is a solution to the problem utt + uxx =0 u(x. Show that the function u (x. 0) − ut (x. u(x. t)| : x ∈ R} = 2 sup{ sinh = 2 sin : x ∈ R} sinh t . t) = 0 is a solution to this problem. t)| : x ∈ R}.

t) − u(x. a small change in the initial data leads to a catastrophically change in the solution. Hence.7 SOLUTIONS AND RELATED TOPICS (c) We have lim sup{|u (x. Thus. the given problem is ill-posed . t)| : x ∈ R} = lim 2 59 t→∞ t→∞ sinh t = ∞.

Exercise 7. dx (b) ∂u = 0 where u = u(x. 0) = sin x Exercise 7.6 Solve each of the following diﬀerential equations: 2 (a) d u = 0 where u = u(x). x) = X(x)T (t). . dx2 ∂2u (b) ∂x∂y = 0 where u = u(x.1 Determine a and b so that u(x. y > 1 u(1. 1) = e. y) = e−2y sin (x − y) is the solution to the initial value problem ux + uy + 2u = 0 for x.4 Show that u(x. y).2 Consider the following diﬀerential equation tuxx − ut = 0.3 Consider the initial value problem xux + (x + 1)yuy = 0. ∂x Exercise 7. Show that there is a constant λ such that X = λX and T = λtT. y > 1 u(x. y) = xex y is the solution to this problem.60 INTRODUCTION TO PDES Practice Problems Exercise 7. Exercise 7. x. Show that u(x.5 Solve each of the following diﬀerential equations: (a) du = 0 where u = u(x). Suppose u(t. y). y) = eax+by is a solution to the equation uxxxx + uyyyy + 2uxxyy = 0. Exercise 7.

y) = f (y + 2x) + xg(y + 2x). where f and g are arbitrary twice diﬀerentiable functions in one variable. Prove that ut = p(u)ux has a solution satisfying u(x. t) = f (x−ct)+g(x+ct). Hint: See Exercise 6.7 SOLUTIONS AND RELATED TOPICS 61 Exercise 7. where f is an arbitrary diﬀerentiable function. Then ﬁnd the general solution to ut = (sin u)ux .2.7 Show that u(x. Exercise 7. .10 Find the general solution to the pde uxx + 2uxy + uyy = 0. Exercise 7. where f and g are two arbitrary twice diﬀerentiable functions.9 Let p : R → R be a diﬀerentiable function in one variable. satisfy the equation uxx − 4uxy + 4uyy = 0.8 Find the diﬀerential equation whose general solution is given by u(x. Exercise 7. t) = f (x + p(u)t).

(a) ux = 3x2 + y 2 . t > 0 sin nx u(x. 0) = 1. t) = 0 for all t ∈ R. t) be a function such that uxx exists and u(0. t) = u(L.62 INTRODUCTION TO PDES Sample Exam Questions Exercise 7. Exercise 7.12 Consider the initial value problem ut + uxx = 0. t > 0 u(x. t)dx ≤ 0. u = u(x. Prove that L uxx (x. (e) Show that the problem is ill-posed. u = u(x. 0) = 1 + . y). t). y) = y − 2x and w(x. n (c) Find sup{|un (x. n2 t (b) Show that un (x. x ∈ R.14 Consider the second-order PDE uxx + 4uxy + 4uyy = 0. t) − 1| : x ∈ R}.11 Let u(x. (a) Show that u(x. z). (d) uxtt = e2x+3t . (d) Find sup{|un (x. 0 Exercise 7. u = u(x.13 Find the general solution of each of the following PDEs by means of direct integration. y. u = u(x. 0) − 1| : x ∈ R}. (c) uxyz = 0. y). t)u(x. x ∈ R. t) = 1 + e n sin nx is a solution to the initial value problem ut + uxx = 0. (a) Use the change of variables v(x. Exercise 7. . (b) uxy = x2 y. y) = x to show that uww = 0. t) ≡ 1 is a solution to this problem. (b) Find the general solution to the given PDE.

7 SOLUTIONS AND RELATED TOPICS Exercise 7.15 Derive the general solution to the PDE utt = c2 uxx by using the change of variables v = x + ct and w = x − ct. 63 .

64 INTRODUCTION TO PDES .

glacier motion. In this chapter we introduce the basic deﬁnitions of ﬁrst order partial differential equations. From a mathematical point of view. ﬁrst order partial diﬀerential equations are most commonly used to describe dynamical processes. and engineering sciences deal with the formulation and the solution of ﬁrst order partial diﬀerential equations. is analyzed. ﬂood waves.First Order Partial Diﬀerential Equations Many problems in the mathematical. and so time. We then derive the one dimensional spatial transport eqution and discuss some methods of solutions. is one of the independent variables. traﬃc ﬂow. First order partial diﬀerential equations and systems model a wide variety of wave phenomena. 8 Classiﬁcation of First Order PDEs In this section. acoustics. Our ﬁrst task is to understand simple ﬁrst order equations. By a ﬁrst order diﬀerential equation in two variables x and y we mean 65 . known as the method of characteristics. physical. One general method of solvability for quasilinear ﬁrst order partial diﬀerential equation. gas dynamics. In applications. we present the basic deﬁnitions pertained to ﬁrst order PDE. ﬁrst order partial diﬀerential equations have the advantage of providing conceptual basis that can be utilized in the study of higher order partial diﬀerential equations. including transport of solvents in ﬂuids. and also a variety of biological and ecological systems. Most of our discussion will focus on dynamical models in a single space dimension. bearing in mind that most of the methods can be readily extended to higher dimensional situations. t.

Examples of nonlinear equations are: ux + cu2 = xy y u2 + u2 = c. If Equation (8.1) can be written in the form a(x. A ﬁrst order pde that is not linear is said to be nonlinear. y. Examples of linear equations are: xux + yuy = cu (y − z)yx + (z − x)uy + (x − y)uz = 0. The following are examples of quasilinear equations: uux + uy + cu2 = 0 x(y 2 + u)ux − y(x2 + u)uy = (x2 − y 2 )u. If Equation (8. y. If Equation (8. u) (8. x y First order partial diﬀerential equations are classiﬁed as either linear or nonlinear. y)uy = c(x. u) (8. linear equations are a special kind of quasilinear equation .3) then we say that the equation is semilinear. The following are examples of semilinear equations: xux + yuy = u2 + x2 (x + 1)2 ux + (y − 1)2 uy = (x + y)u2 .1) can be written in the form a(x.66 FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS any equation of the form F (x. b.2) then we say that the equation is quasilinear.1) In what follows the functions a. u)ux + b(x. u. ux . y) (8. y)ux + b(x. y)ux + b(x. u)uy = c(x. y.1) can be written in the form a(x. uy ) = 0. (8. Clearly. y)u = d(x. y. and c are assumed to be continuously diﬀerentiable functions.4) then we say that the equation is linear. y)uy + c(x. y.

Let v(t. Hence. x) + bux (t. Solution. aut (t. y) = 0. Using the chain rule we see that ut (t. x) = bt − ax.8 CLASSIFICATION OF FIRST ORDER PDES 67 (8. Solution. x) = abfv (v) − abfv (v) = 0 . Recall that for an ordinary linear diﬀerential equation. Likewise. Example 8. x) + bux (t.1 Solve the equation ut (t. x) = f (bt − ax) is a solution to the given equation. x) = 0. x) = bfv (v) and ux (t. the general solution depends on arbitrary functions.4) reduces to a linear equation if c is linear in u. Also. The general solution is given by u(t. Examples of linear homogeneous equations are: xux + yuy = cu (y − z)yx + (z − x)uy + (x − y)uz = 0. Unlike ODEs. x) = −afv (v). semilinear equations (8.2 Consider the transport equation aut (t.2) if a and b are functions of x and y only and c is a linear function of u. Examples of nonhomogeneous equations are: ux + (x + y)uy − u = ex yux + xuy = xy. x) = f (x) where f is an arbitrary differentiable function of x Example 8. the general solution depends mainly on arbitrary constants. y) ≡ 0 and nonhomogeneous if d(x. Show that u(t. A linear equation is called homogeneous if d(x. in linear partial differential equations. semilinear equations are quasilinear equations if a and b are functions of x and y only. where f is an arbitrary diﬀerentiable function in one variable. x) = 0 where a and b are constants.

y) = y − 2 (x2 − y 2 ) satisﬁes the equation 1 1 1 ux + uy = x y y 1 subject to u(x. y) = ex f (2x − y). Exercise 8.2 Show that u(x. (a) xux + yuy = sin (xy). semi-linear. is a solution to the equation ux + 2uy − u = 0. Exercise 8.4 Show that u(x. Exercise 8.68 FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS Practice Problems Exercise 8. y) = x xy satisﬁes the equation xux − yuy = u subject to u(y. x y (d) (x + 3)ux + xy 2 uy = u3 . y) = y 2 . where f is a diﬀerentiable function of one variable.1 Classify each of the following PDE as linear. or nonlinear. quasilinear. y) = cos y 2 . y) = cos (x2 + y 2 ) satisﬁes the equation −yux + xuy = 0 subject to u(0.5 1 Show that u(x. Exercise 8. 1) = 2 (3 − x2 ). (b) ut + uux = 0 (c) u2 + u3 u4 = 0.3 √ Show that u(x. .

and w is a solution to the inhomogeneous equation.6 Find a relationship between a and b if u(x. Suppose v is a solution to the homogeneous equation.8 CLASSIFICATION OF FIRST ORDER PDES 69 Exercise 8. 0) = f (x) ≥ 0. Show u = av + w is a solution to the inhomogeneous equation for any constant a. Suppose that x→∞ lim xf (x) = 0. We will see later in the book that the solution to this problem is a nonnegative ∞ function. . Exercise 8. Show that I(t) is a decreasing function of t. Exercise 8.9 Solve the partial diﬀerential equation ux + uy = 1 by introducing the change of variables s = x + y and t = x − y. L(αu+βv) = αL(u)+βL(v). y) = f (ax + by) is a solution to the equation 3ux − 7uy = 0 for any diﬀerentiable function f.8 Reduce the partial diﬀerential equation aux + buy + cu = 0 to a ﬁrst order ODE by introducing the change of variables s = ax + by and t = bx − ay.7 Suppose L is a linear operator. u(x. Consider the homogeneous and inhomogeneous linear equations Lu = 0 Lu = f where f is some function.10 Consider the IVP −xux + ut = 0. that is. Exercise 8. Deﬁne I(t) = −∞ u(x. t)dx. Exercise 8.

w = t. . a. y) = ax + by and t(x. Write the diﬀerential equation with unknown function w(v).11 Show that u(x. Exercise 8. b = 0 by using the change of variables s(x. y) = e−4x f (2x − 3y) is a solution to the ﬁrst-order PDE 3ux + 2uy + 12u = 0.12 Derive the general solution of the PDE aut + bux = u. Exercise 8. y) in the coordinates v = x − ct. Assume a2 + b2 > 0.15 Suppose that u(x.70 FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS Sample Exam Questions Exercise 8. a.13 Derive the general solution of the PDE aut + bux = 0. Let v = x − ct. Exercise 8.14 Write the equation ut + cux + λu = f (x. y) = bx − ay. Exercise 8. b = 0 1 by using the change of variables v = ax − bt and w = a t. t) = w(x − ct) is a solution to the PDE xux + tut = Au where A and c are constants.

amount of chemical/area) at position x at time t.9 THE ONE DIMENSIONAL SPATIAL TRANSPORT EQUATIONS 71 9 The One Dimensional Spatial Transport Equations Modeling is the process of writing a diﬀerential equation to describe a physical situation.1 Let u(x. Then at time t.) We identify G with the open interval (a. the amount of chemical stored in a section of the tube between positions a and x is given by the deﬁnite integral x Au(s. (The very same discussion applies to the description of the transport of gas by air moving through a pipe. t)ds. and the velocity c > 0 is in the (rightward) positive direction of the x−axis.1 Figure 9. a .e. b). See Figure 9. Linear Transport Equation for Fluid Flows We shall describe the transport of a dissolved chemical by water that is traveling with uniform velocity c through a long thin tube G with uniform cross section A. In this section we discuss the one-dimensional transport equation and discuss an analytical method for solving it. We will assume that the concentration of the chemical is constant across the cross section A at each point x so that the chemical changes in the x−direction and thus the term one-dimensional spatial equation. t) be a continuously diﬀerentiable function denoting the concentration of the chemical (i.

This equation is called the transport equation in one-dimensional space. Example 9. Hence. Solution. since the graph of the function f (x − ct) at a given time t is the graph of f (x) shifted to the right by the value ct. the function f (x) is moving without changes to the right at the speed c An initial value condition determines a unique solution to the transport equation as stated in the next theorem. t) = u(x + ch. Taking the limit of this last equation as h approaches 0 we ﬁnd ut (x. Using the chain rule we ﬁnd ut = −cf (x − ct) and ux = f (x − ct). t) = f (x − ct) is called the right traveling wave. by substituting these results into the equation we ﬁnd ut + cux = −cf (x − ct) + cf (x − ct) = 0 The solution u(x. t)ds = a a+ch u(s.72 FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS Since the water is ﬂowing at a speed c. t).1) for all (x. so at time h + t the same quantity of chemical will be x x+ch Au(s.1). t) + cux (x. t) = 0 (9. t + h)ds. homogeneous ﬁrst order partial diﬀerential equation. t) = f (x − ct) is a solution to (9. It is a linear. t + h). Thus. where f is an arbitrary diﬀerentiable function in one variable. Taking the derivative of both sides with respec to to x we ﬁnd u(x. Now taking the derivative of this last equation we respect to h we ﬁnd 0 = ut (x + ch. t + h). t + h) + cux (x + h. with growing time.1 Show that u(x. .

1) by solving the more general equation aux + buy = 0 where a2 + b2 > 0. t) = 0 u(x.4 Let f be a continuously diﬀerentiable function. Indeed. in terms of x and y we ﬁnd u(x. 0) = f (x).2) . t) to the IVP aut (x. Now. t = bx − ay According to the chain rule for the derivative of a composite function. (9.2) to obtain a2 us + abut + b2 us − abut = 0 or (a2 + b2 )us = 0 and since a2 + b2 > 0 we obtain us = 0. we have ux =us sx + ut tx = aus + but uy =us sy + ut ty = bus − aut Substituting these into (9. t) = f (bt − ax). Solving this equation. Then there is a unique continuously diﬀerentiable solution u(x. t) = f (w) where f is an arbitrary diﬀerentiable function of one variable. y) = f (bx − ay). t) + bux (x. we ﬁnd u(s.9 THE ONE DIMENSIONAL SPATIAL TRANSPORT EQUATIONS 73 Theorem 9. u is given explicitly by the formula u(x. Method of Solutions: The Coordinate Method We will solve (9. We introduce a new rectangular system by the substitution s = ax + by.

u(v. We illustrate this approach in the next example. 0) = e−x . 2 u(x. we introduce the characteristic coordinates given by v = x − ct. we ﬁnd e−x = f (x). Hence.74 FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS Example 9. Hence. Then ux = −3uv + uw and ut = uv + 3uw . Substituting these into the given equation we ﬁnd 10uv = 0 or uv = 0. t) where λ and c are constants and f is a given function representing external resources. Example 9. w) which can be solved by the method of integrating factor. dt A transport equation with growth is an equation given by ut + cux + λu = f (x.2 2 Use the coordinate method to ﬁnd the solution to ut −3ux = 0. t) = f (x + 3t) where f is a diﬀerentiable 2 2 function in one variable. Solution. 0) = e−x . Since u(x. .3 Find the general solution of the transport equation ut + ux − u = t. To solve this equation. Note that the growth is characterized by the term λu. Let v = −3x + t and w = x + 3t. we ﬁnd ut =uv vt + uw wt = −cuv + uw ux =uv vx + uw wx = uv Substituting these into the original equation we obtain the equation uw + λu = f (v + cw. t) = e−(x+3t) Transport Equation with Decay: The Method of Characteristic Coordinates Recall from ODE that a function u is an exponential growth function if it satisﬁes the equation du = λu. w) = f (w) or u(x. w = t. Using the chain rule. u(x.

w = t. we ﬁnd d −w (e u) = we−w dw and solving this equation we ﬁnd u(v. These transform the original equation to the ﬁrst order ODE uw − u = w. known as the method of characteristics. t) = f (x − t)et − (1 + t) 75 A more general method for solving quasilinear ﬁrst order partial diﬀerential equations.9 THE ONE DIMENSIONAL SPATIAL TRANSPORT EQUATIONS Solution. . will be discussed in the next section. w) = −(1 + w) + ew f (v) and in terms of x and y we ﬁnd u(x. The characteristic coordinates are v = x − t. Using the method of integrating factor.

1+x2 Exercise 9.7 Show that the decay term λu in the transport equation with decay ut + cux + λu = 0 can be eliminated by the substitution w = ueλt .6 Solve ut + ux − 3u = t with initial condition u(x. u(x. 0) = x2 . 0) = 1 . where f : R → R is a given function. Exercise 9. solve the equation ux + 2uy = cos (y − 2x) with the initial condition u(0.8 Use the coordinate method to solve ux + uy = u2 u(x.76 FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS Practice Problems Exercise 9. solve the equation aux + buy + cu = 0.1 Use the coordinate method to ﬁnd the solution to ut + 3ux = 0. 0) = sin x. 0) = h(x) . Exercise 9. y) = f (y). Exercise 9.4 Show that the initial value problem ut + ux = x.3 Use the coordinate method. Exercise 9. u(x. Exercise 9. Exercise 9. x) = 1 has no solution.5 Solve the transport equation ut + 2ux = −3u with initial condition u(x.2 Use the coordinate method.

t) = u(x. 0) = f (x) − g(x) 77 (b) Write an explicit formula for w in terms of f and g.10 Solve the initial boundary value problem ut + cux = −λu. t) is the unique solution to the IVP ut + cux = 0 u(x. That is. t > 0 u(x.9 (Well-Posed) Let u be the unique solution to the IVP ut + cux = 0 u(x. x > 0. 0) = f (x) and v be the unique solution to the IVP ut + cux = 0 u(x. t) − v(x. t > 0. Exercise 9. u(0. 0) = g(x) where f and g are continuously diﬀerentiable functions. a small change in the initial data leads to a small change in the solution.9 THE ONE DIMENSIONAL SPATIAL TRANSPORT EQUATIONS Exercise 9. t) = g(t). (c) Use (b) to conclude that the transport problem is well-posed. 0) = 0. (a) Show that w(x. .

B.12 Solve the PDE ux + uy = 1 using the coordinate method. A fairly general method iin ﬁnding a particular solution is to use an integrating factor. B. (a) Determine a. Exercise 9. and C are constants with A = 0 and G(x. by writing the Auv +Cu = G(x.11 Solve the ﬁrst-order equation 2ut +3ux = 0 with the initial condition u(x. y) an arbitrary function. Exercise 9. c. y) where A. the general solution to this equation is the sum of the general solution of the homogeneous equation and a particular solution of the nonhomogeneous equation. C such that ad − bc = 0 and so that the change of variables v = ax + by and w = cx + dy will reduce the give PDE to a ﬁrst order PDE of the form αuv + βu = 0. Exercise 9. (Bv − w) . ∂v A . exactly as with ODEs.15 Consider the ﬁrst order linear nonhomogeneous PDE Aux + Buy + Cu = G(x. b. y) based on the formulas v = x and w = Bx−Ay in the form C C 1 ∂ (Aue A v ) = e A v G v.78 FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS Sample Exam Questions Exercise 9. and C are constants with A = 0.14 Use the result of the previous section to solve the PDE ux + uy + u = 0. d in terms of A. As in the case of ODEs. Exercise 9. B.13 Consider the ﬁrst order linear homogeneous PDE Aux + Buy + Cu = 0 where A. 0) = sin x. (b) Use (a) to ﬁnd the general solution of the given PDE.

Using the above idea.9 THE ONE DIMENSIONAL SPATIAL TRANSPORT EQUATIONS 79 The right-hand side may be integrated with respect to v. . treating w as a constant. ﬁnd the general solution of the PDE ux + uy + u = x + y.

1 Sometimes (10. a b c (10. y. u) and w(x. u)ux + b(x. u) is given by f (v.80 FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS 10 The Method of Characteristics In this section we develop a method for ﬁnding the general solution of a quasilinear ﬁrst order partial diﬀerential equation. u)uy = c(x.1) Equations (10. This method of solution can be described by the following result. w =constant= c2 are solutions to the ODE system dy du dx = = .3) (10.1 Find the general solution of the pde x2 ux + y 2 uy = (x + y)u. This method is called the method of characteristics or Lagrange’s method. w) = 0 (10.2) is written explicitly as v = g(w) or w = g(v) where g is an arbitrary diﬀerentiable function. y. The corresponding parametric forms are given by the system of ODEs dx =a ds dy =b ds du =c ds Remark 10. Solution. (x+y)u = = Using the ﬁrst . Theorem 10. y. y. u) and v =constant= c1 .3) are called the characteristic equations in non-parametric forms. Example 10. The characteristic equations for this pde are dx x2 dy y2 du .1 The general solution of the quasilinear ﬁrst order PDE a(x.2) where f is an arbitrary diﬀerentiable function of v(x. y.

we have solve for x obtaining x = tions we ﬁnd du u dy y2 dx = dy and x2 y2 1 and x + y 1 −c1 y 81 this implies x−y = c1 . u2 − y 2 ) = 0 or u2 = y 2 + g(x2 − y 2 ). where f and g are arbitrary diﬀerentiable functions Example 10. Hence. Using the last two fractions we ﬁnd u2 − y 2 = c2 . the general solution is f x − y xy . Solution. Also. the general solution is f (x2 − y 2 .3 Find the general solution of the pde x(y 2 − u2 )ux − y(u2 + x2 )yy = (x2 + y 2 )u. Using the ﬁrst two fractions xy we ﬁnd x2 − y 2 = c1 . (x2 +y 2 )u Using a xdx + ydy + udu du = 2 . Hence. xy u =0 where f is an arbitrary diﬀerentiable function Example 10. we can xy y = 1−c1 y + y. Using the last two fracdu u = du u(x+y =⇒ 1 y 1 y2 y 1−c1 y du u + y dy = 2 y =⇒ = y−c1 y 2 du u + 1 y dy = =⇒ 1−c1 y y−c1 y 2 + c1 1−c1 y + dy = y2 1−c1 y =⇒ + ln (1 − c1 y) = ln z + c2 =⇒ = ec2 z =⇒ y c1 1−c1 y y · 1−c1 y =⇒ 2 ln y − y 1 = c2 z =⇒ y 1 −c1 = c2 z =⇒ xy = c2 z.2 Find the general solution of the pde yuux + xuuy = xy. 2 (u2 + x2 ) + u2 (x2 + y 2 −y (x + y 2 )u xdx + ydy + udu du = 2 0 (x + y 2 )u xdx + ydy + udu = 0. dx The characteristic equations are x(y2 −u2 ) = property of proportions we can write x2 (y 2 That is − u2 ) dy −y(u2 +x2 ) = du . or . Solution. dy dx The characteristic equations are yu = xu = du .10 THE METHOD OF CHARACTERISTICS two fractions.

The last fraction implies u = k .4 Solve the transport equation using the method of characteristics ut + cux = 0. These are curves in the xy−plane. The general solution is given yu =0 x f x2 + y 2 + u2 .82 FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS Hence. we ﬁnd ln by or =constant or yu x = c2 . Example 10. x y u yu x Hence. u) = 0 or u = g(x − ct) dy b = dx a are called characteristic curves or simply characteristics. Also. Solution. 1 c 0 Solving the ﬁrst two fractions we ﬁnd x − ct = k.5 Find the characteristics of cos yux + uy + xu = 0. The characteristic equations are given by dx du dt = = . u= x g(x2 + y 2 + u2 ) y where f and g are arbitrary diﬀerentiable functions Example 10. The general solution is given by f (x − ct. dx x u2 − + y2 or − dy y u2 + x2 = (x2 du + y 2 )u du dx dy − = . Solution curves to the ODE . we ﬁnd x2 + y 2 + u2 = c1 .

Thus. We have a = 1 and b = 2.6 Find the characteristics of ux + 2uy − u = 0.10 THE METHOD OF CHARACTERISTICS Solution. Solving the equation sin y − x = k 83 dy dx = 1 cos y by the separation of variable method we ﬁnd Example 10. by 2x − y = k dy dx = 2 so that the characteristics are given . Solution.

3 Find the characteristics of the pde (x + y)(ux + uy ) = u − 1 Exercise 10.4 Find the general solution of the pde xux + yuy = 1 + u2 . Exercise 10. Exercise 10.1 Find the characteristics of the pde xux − yuy = u Exercise 10. Exercise 10.10 Find the general solution of the pde (y + xu)ux − (x + yu)uy = x2 − y 2 .9 Find the general solution of the pde u(u2 + xy)(xux − yuy ) = x4 .6 Find the general solution of the pde xux + yuy = u. Exercise 10.7 Find the general solution of the pde xux + yuy = nu.11 Find the general solution of the pde (y 2 + u2 )ux − xyuy + xu = 0. Exercise 10. Exercise 10.8 Find the general solution of the pde x(y − u)ux + y(u − x)uy = u(x − y).2 Find the characteristics of the pde −yux + xuy = 0 Exercise 10.84 FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS Practice Problem Exercise 10. Exercise 10. .5 Find the general solution of the pde ln (y + u)ux + uy = −1.

13 Find the general form of solutions to (1 + x2 )ux + uy = 0 and sketch some of the characteristics. .10 THE METHOD OF CHARACTERISTICS Exercise 10.12 Find the general form of solutions to ux + 2uy = u 85 and sketch some of the characteristics. Hint: deﬁne a new variable v = e−x u. What equation does v satisfy? Exercise 10.

18 Find the characteristics of the pde 1 1 1 ux + uy = x y y . Exercise 10. Exercise 10.15 Find the characteristics associated with the PDE ux + xuy + 3u = 2.86 FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS Sample Exam Questions Exercise 10.17 Find the characteristics of the pde 1 1 ux + uy = 0 x y Exercise 10.14 Find the general solution of the equation ux + yuy = u. Exercise 10.16 Find the general solution of hte ﬁrst order PDE ux + yuy + xu = 0.

i. y.2) We call C the initial curve of the problem.1) Recall that the initial value problem of a ﬁrst order ordinary diﬀerential equation asks for a solution of the equation which has a given value at a given point in R. In general. we determine the particular solution by formulating an initial value porblem also known as a Cauchy problem. u(x0 (t). Let u0 (t) be a given continuously diﬀerentiable function on I. Initial Value Problem or Cauchy Problem Let C be a given curve in R2 deﬁned parametrically by the equations x = x0 (t).e. y) deﬁned in a domain Ω ⊂ R2 containing the curve C and such that: (1) u = u(x. u)ux + b(x. y0 are continuously diﬀerentiable functions on some interval I. u0 (t) the initial data.1. y = y0 (t) where x0 . (2) On the curve C. y. u)uy = c(x. See Figure 11.1) in Ω. A precise statement of the problem is given next.1) asks for a solution of (11. we discuss the Cauchy problem for the ﬁrst order quasilinear partial diﬀerential equation a(x. In the case of a ﬁrst order partial diﬀerential equation. y) is a solution of (11. u). The Cauchy problem for (11. t ∈ I. one deals with those partial diﬀerential equations whose solutions satisfy certain supplementary conditions. .2) the initial condition of the problem. (11. In this section. and (11. y0 (t)) = u0 (t). y. u equals the given function u0 (t).1) asks for a continuously diﬀerentiable function u = u(x. it is seldom the case that one tries to study the properties of the general solution of such equations.11 THE CAUCHY PROBLEM FOR FIRST ORDER QUASILINEAR EQUATIONS87 11 The Cauchy Problem for First Order Quasilinear Equations When solving a partial diﬀerential equation. The Cauchy problem for the PDE (11.1) which has given values on a given curve in R2 . (11.

and c are functions of x.2) has a unique solution. . and u with continuous ﬁrst order partial derivatives with respect to their argument in some domain D of (x. u0 (t)) dx0 dy0 (t) − b(x0 (t).2.1 If we view a solution u = u(x. y = y0 (t). y.1) . and u0 (t) are continuously diﬀerentiable functions of t in an interval I. u = u0 (t). Then for each point (x0 (t). u = u0 (t) where t ∈ I. Theorem 11. dt dt (11. t ∈ I.1) as an integral surface of (11. y0 (t). u0 (t)) (t) = 0.1). y0 (t). and that a.2) is satisﬁed for every point on C contained in U. y0 (t).1 Suppose that x0 (t).1) containing the curve Γ described parametrically by the equations Γ : x = x0 (t). The following theorem asserts that under certain conditions the Cauchy problem (11. u0 (t)) on Γ that satisﬁes the condition a(x0 (t). u)−space containing the initial curve Γ : x = x0 (t). y. y0 (t)) such that the initial condition (11. y0 (t). y) of (11. y = y0 (t). See Figure 11. b.3) there exists a unique solution u = u(x. Note that the projection of this curve in the xy−plane is just the curve C.1) in a neighborhood U of (x0 (t). we can give a simple geometrical statement of the problem: Find a solution surface of (11. y) of (11.(11.88 FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS Figure 11.

u0 (t)) ∈ Γ. The solution depends on the parameter s so it consists of a triples of functions x = x(s. Using this as the initial value we solve the system of ODEs consisting of the characteristic equations in parametric form dx =a ds dy =b ds du =c ds satisfying the initial condition (x(0). y0 . u0 ). y0 (t).2 Note that condition (11.3) implies that b(x0 . u(0)) = (x0 (t). y(0). u0 ) dy0 (t) = dx0 (t) a(x0 . u0 ). We construct the desired solution using the method of characteristics as follows: Pick a point (x0 (t). u0 ) which means that the vector (a(x0 . y0 . It follows that the Cauchy problem has a unique solution if C is nowhere characteristic. y0 (t). y0 . u0 )) is not tangent to Γ. y = y(s. t). c(x0 . u = u(s. y0 . y0 . u0 (t)). b(x0 . t) (11.4) .11 THE CAUCHY PROBLEM FOR FIRST ORDER QUASILINEAR EQUATIONS89 Figure 11. t).

y0 (t). y)). To ﬁnd this solution. t) = s. The initial curve in R3 can be given parametrically as Γ : x0 (t) = t. y) = u(s(x. y). t) = t + s. ds Solving this system we ﬁnd x(s) = s + c1 . y). y(s) = s + c2 . u(s) = s + c3 . u0 (t)) dx0 dy0 (t) − b(x0 (t). t) = s + f (t). s = s(x. Similarly. u0 (t)) (t) = −1 = 0 dt dt so by the above theorem the given Cauchy problem has a unique solution. y) and substituting these into the third equation to obtain u(x. y0 (t) = 0.1 Solve the Cauchy problem ux + uy =1 u(x. t(x. u0 (t) = f (t). 0) =f (x) Solution.4) for t = t(x. The solution u is recovered by solving the ﬁrst two equations in (11. we solve the system of ODEs dx =1 ds dy =1 ds du =1. y0 (t).90 FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS This system represents the parametric representation of the integral surface of the problem in which the curve Γ corresponds to s = 0. y(s. We have a(x0 (t). . Example 11. But x(0) = t so that c1 = t. the unique solution is given parametrically by the equations x(s. Hence. u(s. y(0) = 0 so that c2 = 0 and u(0) = f (t) implies c3 = f (t).

we obtain inﬁnitely many solutions. Pick a new initial curve Γ passing through P0 which is not tangent to Γ at P0 . . The initial curve in R3 can be given parametrically as Γ : x0 (t) = t.3) is not satisﬁed than C is a characteristic curve. Using the Cauchy data u(x. Alternative Computation We can apply the results of the previous section to ﬁnd the unique solution. condition (11. the unique solution is given by u(x. y) = x + f (x − y) − (x − y) = y + f (x − y) If condition (11. 0) = f (x) we ﬁnd f (x) = x + F (x) which implies that F (x) = f (x) − x. Thus. the general solution of the PDE is given by u = x + F (x − y). Hence. We illustrate this case in the next example.3) is satisﬁed and the new Cauchy problem has a unique solution.11 THE CAUCHY PROBLEM FOR FIRST ORDER QUASILINEAR EQUATIONS91 Solving the ﬁrst two equations for s and t we ﬁnd s = y. If the curve Γ satisﬁes the characteristic equations than the problem has inﬁnitely many solutions. y0 . u0 ) on Γ. t = x − y and substituting these into the third equation we ﬁnd u(x.2 Solve the Cauchy problem ux + uy =1 u(x. In this case. Example 11. y) = y + f (x − y). To see this. x) =x Solution. Since there are inﬁnitely many ways of selecting Γ . If we solve the characteristic equations in non-parametric form dx dy du = = 1 1 1 we ﬁnd x − y = c1 and u − x = c2 . y0 (t) = t. u0 (t) = t. pick an arbitrary point P0 = (x0 .

Using the Cauchy data u(x. x) = x we ﬁnd f (0) = 0. y0 (t).3 Solve the Cauchy problem ux + uy =1 u(x. Thus. dt dt Solving the characteristic equations in parametric form we ﬁnd x(s) = s + c1 . which is not possible since the LHS is a ﬁxed number whereas the RHS is a variable expression. x) =1 Solution. There are inﬁnitely many choices for f. y(s) = s + c2 . y0 (t) = t. the solution is given by u(x. the problem has inﬁnitely many solutions. dt dt As in Example 11. Γ does not satisfy the characteristic equations. Example 11. Note that Γ satisﬁes the characteristic equations If condition (11. u(s) = s + c3 .1. Clearly. the problem has no solutions . u0 (t) = 1. y0 (t). the general solution to the PDE is given by u = x + f (x − y). the general solution of the PDE is u = x+f (x−y) where f is an arbitrary diﬀerentiable function. We have a(x0 (t). y0 (t). We illustrate this case next. Hence.92 We have FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS a(x0 (t). u0 (t)) (t) = 0. x) = 1 we ﬁnd f (0) = 1 − x. u0 (t)) (t) = 0. Hence. u0 (t)) dy0 dx0 (t) − b(x0 (t).3) is not satisﬁed and if Γ does not satisfy the characteristic equations then it can be shown that the Cauchy problem has no solutions. y) = y + f (x − y) where f is an arbitrary function such that f (0) = 0. Using the Cauchy data u(x. Now. The initial curve in R3 can be given parametrically as Γ : x0 (t) = t. y0 (t). u0 (t)) dx0 dy0 (t) − b(x0 (t).

u0 (t)) (t) = 1 = 0 dt dt (11. u0 (s) = 1. The initial curve is given parametrically by Γ : x0 (s) = s. u0 (t) = t2 . y0 (s) = 0. . y0 (t). 0) = 1. the general solution can be represented by u − x = f (x + y) where f is an arbitrary diﬀerentiable function. 1 −1 1 Using the ﬁrst two fractions we ﬁnd x + y = c1 .11 THE CAUCHY PROBLEM FOR FIRST ORDER QUASILINEAR EQUATIONS93 Example 11. The characteristic equations are dy du dx = = . We have a(x0 (t).4 Solve the Cauchy problem ux − uy =1 u(x. Thus. Using the ﬁrst and the third fractions we ﬁnd u − x = c2 .5 Solve the initial value problem ut + uux = x. Hence. the unique solution is given by u = x + (x + y)2 − (x + y) = (x + y)2 − y Example 11.5) so the Cauchy problem has a unique solution. 0) = x2 we ﬁnd x2 − x = f (x). The initial curve is given parametrically by Γ : x0 (t) = t. y0 (t). 0) =x2 Solution. Solution. Using the Cauchy data u(x. u(x. u0 (t)) dy0 dx0 (t) − b(x0 (t). y0 (t) = 0.

Thus. s > 0 ds ds . This can be reduced further as follows: u + ue−2t = x + 2e−t − xe−2t = −2t 2e−t 2e−t + x(1 − e−2t ) =⇒ u = 1+e−2t + x 1−e−2t = sech(t) + xtanh(t) 1+e Example 11. Hence. u0 (s)) (s) = −1 = 0 ds ds so the Cauchy problem has a unique solution. u2 − x2 ) = 0 where f is an arbitrary diﬀerentiable function. The characteristic equations are dx du dt = = . u0 (s)) dy0 dx0 (s) − b(x0 (s). Now. Using the Cauchy data we ﬁnd c1 = 1 + x and c2 = 1 − x2 = 2(1 + x) − (1 + x)2 = 2c1 − c2 . u0 (s)) (s) = −4s = 0. the general solution is givne by f ((x + u)e−t . We have a(x0 (s). y0 (s). 1 u2 − x2 = 2(x + u)e−t − (x + u)2 e−2t or u − x = 2e−t − (x + u)e−2t . 1 u x Since dt d(x + u) = 1 x + u) we ﬁnd that t = ln (x + u) − ln c1 or (x + u)e−t = c1 . y0 (s) = 2s.94 We have FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS a(x0 (s). y0 (s). u0 (s) = 0. Solution. s > 0. y0 (s). y0 (s).6 Solve the initial value problem uux + uy = 1 with the initial curve Γ : x0 (s) = 2s2 . using the last two fractions we ﬁnd u2 − x2 = c2 . u0 (s)) dy0 dx0 (s) − b(x0 (s).

Solving this quadratic equation in u to ﬁnd 2u = y ± (4x − y 2 ) 2 . t) = t + c2 . 2 Using the initial conditions x(s. 0) = 2s2 . This represents a solution surface only when y 2 < 4x. u(s. Solving the ﬁrst equation with u being replaced by t + c3 we ﬁnd 1 x(s. 2 Eliminating s and t we ﬁnd (u − y)2 + u2 = 2x. the solution of this system depends on two parameters s and t. t) = t. y(s. The solution does not exist for y 2 > 4x 1 1 we ﬁnd . 0) = 2s. The characteristic equations in parametric form are given by the system of ODEs dx =u dt dy =1 dt du =1 dt Thus.11 THE CAUCHY PROBLEM FOR FIRST ORDER QUASILINEAR EQUATIONS95 so the Cauchy problem has a unique solution. 0) = 0 1 x(s. t) = t + c3 . u(s. u(s. t) = t + 2s. The solution surface satisfying u = 0 on y 2 = 2x is given by 2u = y − (4x − y 2 ) 2 . t) = t2 + c3 t + c1 . Solving the last two equations we ﬁnd y(s. t) = t2 + 2s2 . y(s.

x ) = 0. Exercise 11. u(x.4 Solve xux + yuy = xe−u with the Cauchy data u(x.5 Solve the initial value problem xux + uy = 0. Exercise 11. Exercise 11.96 FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS Practice Problems Exercise 11. 0) = x3 .3 Solve x(y 2 + u)ux − y(x2 + u)uy = (x2 − y 2 )u with the Cauchy data u(x. u(x. 0) = cos x . Exercise 11.7 Solve the initial value problem aux + uy = u2 .6 Solve the initial value problem ut + aux = 0. x2 ) = 0.2 Solve the linear equation yux + xuy = u with the Cauchy data u(x. u(x. 0) = f (x) using the characteristic equations in parametric form.1 Solve (y − u)ux + (u − x)uy = x − y 1 with the condition u(x. 0) = f (x). Exercise 11. Exercise 11. −x) = 1.

u(x.10 Solve the initial value problem √ 1 − x2 ux + uy = 0.9 Solve the initial value problem uux + uy = 0.11 THE CAUCHY PROBLEM FOR FIRST ORDER QUASILINEAR EQUATIONS97 Exercise 11.8 Solve the initial value problem ux + xuy = u. 0) = f (x) Exercise 11. u(1. u(x. y) = h(y) Exercise 11. 0) = sin x .

(i) Find and sketch the characteristics. .14 Solve the Cauchy problem ux + 4uy = x(u + 1) u(x. y) = ey . using your picture of the characteristics.13 (a) Find the general solution of the equation ux + yuy = u.98 FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS Sample Exam Questions Exercise 11. 0) = h(x) for given functions g and h? (iv) Explain.12 Solve the equation ux + uy = u subject to the condition u(x. 0). (b) Find the solution satisfying the Cauchy data u(x. (iii) What happens if you try to ﬁnd the solution satisfying either u(0. −x) = sin x. 3ex ) = 2. Exercise 11. (c) Find the solution satisfying the Cauchy data u(x.11 Consider xux + 2yuy = 0. Exercise 11. y) = (0. what goes wrong at (x. 0) = cos x. (ii) Find the solution with u(1. ex ) = ex . 5x) = 1. y) = g(y) or u(x. Exercise 11.15 Solve the Cauchy problem ux − uy = u u(x. Exercise 11.

0) = 1 (b) u(x. (d) In which region of the plane is the solution uniquely determined? Exercise 11. 2 (c) Find the solution satisfying the boundary condition u(0. if no.17 Consider the equation ux + yuy = 0. (b) Sketch some of the characteristics.16 (a) Find the characteristics of the equation yux + xuy = 0. explain why. give a formula. . 0) = x? If yes.11 THE CAUCHY PROBLEM FOR FIRST ORDER QUASILINEAR EQUATIONS99 Exercise 11. y) = e−y . Is there a solution satisfying the extra condition (a) u(x.

100 FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS .

and elliptic type. These types arise in many applications such as the wave equation.Second Order Linear Partial Diﬀerential Equations In this chapter we consider the three fundamental second order linear partial diﬀerential equations of parabolic.4) 101 (12. u. ux . uy )uyy = D(x.1) can be written in the form A(x. u. y) (12. u. uxy ) = 0. y)uxy + C(x. uy ) (12. ux . ux . y)uyy = D(x. If Equation (12. y)ux +E(x. 12 Second Order PDEs in Two Variables In this section we will brieﬂy review second order partial diﬀerential equations. y)uyy +D(x. u.3) .1) can be written in the form A(x. ux .2) (12. uyy . uy )uxx +B(x. uy )uxy +C(x. y)uxx +B(x. uxx . the heat equation and the Laplace’s equation. uy .1) can be written in the form A(x. y. y. y. y)u = G(x. hyperbolic. ux . ux . y. u. y)uxx + B(x. We will study the solvability of each of these equations. If Equation (12. y. uy ) then we say that the equation is semilinear. y. A second order partial diﬀerential equation in the variables x and y is an equation of the form F (x. If Equation (12. y)uy +F (x. u.1) then we say that the equation is quasilinear. y)uxy +C(x.

The Laplace equation is given by ∆u = uxx + uyy = 0. The wave equation in one-dimensional space is given by utt = c2 uxx where c is a constant. Example 12.4) resembles the general equation of a conic section Ax2 + Bxy + Cy 2 + Dx + Ey + F = 0 which is classiﬁed as either parabolic. Equation (12. parabolic or elliptic. • Elliptic: This occurs if B 2 − 4AC < 0 at a given point in the domain of u. B = 0. and C = x. y) = 0 and nonhomogeneous otherwise. 2. or elliptic based on the sign of the discriminant B 2 − 4AC. Solution. Since B 2 − 4AC = −4x. In what follows we list some of the well-known models that are of great interest: 1. The heat equation in one-dimensional space is given by ut = kuxx where k is a constant. Here we are given A = 1. We do the same for a second order linear partial diﬀerential equation: • Hyperbolic: This occurs if B 2 − 4AC > 0 at a given point in the domain of u. • Parabolic: This occurs if B 2 − 4AC = 0 at a given point in the domain of u. A linear equation is said to be homogeneous when G(x.1 Determine whether the equation uxx + xuyy = 0 is hyperbolic. parabolic if x = 0 and elliptic if x > 0 Second order partial diﬀerential equations arise in many areas of scientiﬁc applications. 3. the given equation is hyperbolic if x < 0. hyperbolic. .102SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS then we say that the equation is linear.

Exercise 12. (a) ut + uux = vuxx (b) xutt + tuyy + u3 u2 = t + 1 x (c) utt = c2 uxx (d) u2 + ux = 0. Exercise 12.12 SECOND ORDER PDES IN TWO VARIABLES 103 Practice Problems Exercise 12. parabolic or elliptic.4 Show that u(x.2 Classify the following linear scalar PDE with constant coeﬃcents as hyperbolic. Sketch these regions. Exercise 12. 0) ut (x. tt = = = = uxx 0 cos x 0 .5 Classify each of the following PDE as linear. parabolic. semi-linear. t) = cos x sin t is a solution to the problem utt u(x. or parabolic. t) for all x.3 Find the region(s) in the xy−plane where the equation (1 + x)uxx + 2xyuxy − y 2 uyy = 0 is elliptic.1 Classify each of the following equation as hyperbolic. c > 0 (b) Heat conduction: ut = cuxx . or elliptic: (a) Wave propagation: utt = c2 uxx . Exercise 12. (c) Laplace’s equation: ∆u = uxx + uyy = 0. hyperbolic. quasilinear. 0) ux (0. (a) uxx + 4uxy + 5uyy + ux + 2uy = 0 (b) uxx − 4uxy + 4uyy + 3ux + 4u = 0 (c) uxx + 2uxy − 3uyy + 2ux + 6uy = 0. or nonlinear. t > 0. c > 0.

9 Show that for any n ∈ N. and that. 0. z) = √ uxx + uyy + uzz = 0. y) = (0. Exercise 12. 1 x2 +y 2 +z 2 is a solution of . y) = ln (x2 + y 2 ) is a solution of uxx + uyy = 0. Exercise 12. Exercise 12. where k0 and kL are given positive numbers and u = u(x). 0).104SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS Exercise 12. Exercise 12. L]. where f and g are arbitrary diﬀerentiable functions. u(x.7 Consider the eigenvalue problem uxx = λu. y) = sin nx sinh ny is a solution to the Laplace equation ∆u = uxx + uyy = 0. the function un (x.6 Show that. 0). u(x. is a solution to the PDE uuxy = ux uy . for all (x. y. z) = (0.8 Show that u(x. y) = f (x)g(x). Can this system have a nontrivial solution u ≡ 0 for λ > 0? Hint: Multiply the ﬁrst equation by u and integrate over x ∈ [0. for all (x. y.10 Solve uxy = xy. 0 < x < L ux (0) = k0 u(0) ux (L) = −kL u(L) with Robin boundary conditions.

or elliptic: (a) 2uxx − 4uxy + 7uyy − u = 0 (b) uxx − 2 cos xuxy − sin2 xuyy = 0 (c) yuxx + 2(x − 1)uxy − (y + 2)uyy = 0.11 Classify each of the following second-oder PDEs according to whether they are hyperbolic.12 SECOND ORDER PDES IN TWO VARIABLES 105 Sample Exam Questions Exercise 12. u(x.14 Consider the second-order hyperbolic PDE uxx + 2uxy − 3uyy = 0. Determine the region D in R2 . ux (x. uxx . Use the change of variables v(x.15 Solve the Cauchy problem uxx + 2uxy − 3uyy = 0. (b) parabolic. parabolic. y) = x + y to solve the given equation. 2x) = 1. t) = (f (x + ct) + f (x − ct)) + 2 2c is a solution to the PDE utt = c2 uxx where f is twice diﬀerentiable function and g is a diﬀerentiable function. Exercise 12. Exercise 12. y) = y − 3x and w(x. 2x) = x. if such a region exists.12 Let c > 0. that makes this PDE: (a) hyperbolic. (c) elliptic. By computing ux .13 Consider the second-order PDE yuxx + uxy − x2 uyy − ux − u = 0. 0) and ut (x. ut . and utt show that 1 1 u(x. Then compute and simplify u(x. 0). x+ct g(s)ds x−ct . Exercise 12. Exercise 12.

1. t). ρ. t). one of the simplest systems to visualize and describe are waves on a stretched elastic string.106SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS 13 Hyperbolic Type: The Wave equation The wave equation has many physical applications from sound waves in air to magnetic waves in the Sun’s atmosphere. A displacement of a tiny piece of the string between points P and Q is shown in Figure 13.1 To derive the wave equation we need to make some simplifying assumptions: (1) The density of the string. The aim is to try and determine the vertical displacement from the x−axis of the string. u(x. Then we distort it by displacing it in the vertical direction and at some time. is constant so that the mass of the string between P and Q is simply ρ times the length of the string between P and Q. u(x. namely ∆s where ∆s = (∆x)2 + (∆u)2 = ∆x 1 + ∆u ∆x 2 ≈ ∆x 1 + ∂u ∂x 2 (2) The displacement. However. and its derivatives are assumed small so that ∆s ≈ ∆x . Initially the string is horizontal. say t = 0. as a function of position x and time t. Figure 13. we release it and the string starts to oscillate.

t) cos θ(x + ∆x. t) T (x. these forces must balance and so T (x. which has magnitude T (x + ∆x. t). t) = T (x + ∆x. t) sin θ(x + ∆x. t) sin θ(x. t) sin θ(x. Since there is no horizontal motion. Then Newton’s Law of motion gives mass × acceleration = Applied Forces that is ρ∆x ∂ 2u = T (x + ∆x. There is no horizontal motion of any portion of the string. t). t) cos θ(x + ∆x. t) = T. (4) The motions are purely vertical. These forces are: (i) tension pulling to the right. t) cos θ(x.13 HYPERBOLIC TYPE: THE WAVE EQUATION and the mass of the portion of the string is ρ∆x. 107 (3) The only forces acting on this portion of the string are the tensions T (x. t) whereas at Q the force is T (x + ∆x. t). t) sin θ(x. (ii) tension pulling to the left. t) above the horizontal. tension is the magnitude of the pulling force exerted by a string). t) cos θ(x. ∂t2 Dividing by T we obtain ρ ∂ 2 u T (x + ∆x. t). and acts at an angle θ(x + ∆x. ∆x→0 ∆x tan θ(x. we consider the forces acting on the typical string portion shown in Figure 13. t) above the horizontal. t) = tan θ(x + ∆x. and acts at an angle θ(x. t) at P and T (x + ∆x. t) = lim . Next. t) whereas at Q the force is T (x + ∆x. t) − T (x. t). t). t) ∆x 2 = − T ∂t T (x + ∆x. t) sin θ(x + ∆x. • Horizontal: At P the tension force is T (x. t) at Q. t). t) cos θ(x. t) − tan θ(x. t) T (x. Now we resolve the forces into their horizontal and vertical components. t) sin θ(x + ∆x. But ∆u = ux (x. The gravitational force is neglected. • Vertical: At P the tension force is −T (x. which has magnitude T (x. (In physics.1. t) cos θ(x + ∆x.

T Dividing by ∆x and letting ∆x → 0 we obtain ρ utt (x. t) = ux (x + ∆x. t) = c2 uxx (x. We call c the wave speed. t) T or utt (x. t) = ux (x + ∆x. t) − ux (x. Now. . Then by application of the chain rule we ﬁnd ut ux utt uxx =c(uv − uw ) =uv + uw =c2 (uvv − 2uvw + uww ) =uvv + 2uvw + uww (13. t). ρ D’Alembert Solution of (13. t) = uxx (x. t). t) where c2 = T .1) we obtain c2 (uvv + 2uvw + uww ) = c2 (uvv − 2uvw + uww ) and this simpliﬁes to 4c2 uvw = 0 or uvw = 0. D’Alembert’s solution involves two arbitrary functions that are determined (normally) by two initial conditions. writing u in terms of x and y we ﬁnd the general solution u(x. It follows that u(v. tan θ(x + ∆x.1) Substituting into (13. Hence.108SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS Likewise. y) = f (x + ct) + g(x − ct).1) Let v = x + ct and w = x − ct. we get ρ ∆xutt (x. w) = f (v) + g(w) where f and g are arbitrary diﬀerentiable functions.

13 HYPERBOLIC TYPE: THE WAVE EQUATION Example 13. We have u(x.1 Find the solution to the initial value problem utt =c2 uxx u(x. 0) = f (x) + g(x) = v(x) and ut (x. 1 1 g(x) = (v(x) − W (x)) 2 c Hence. t) = [v(x − ct) + v(x + ct) + 2 1 = [v(x − ct) + v(x + ct) + 2 1 (W (x + ct) − W (x − ct))] c 1 x+ct w(s)ds] c x−ct w(x)dx 109 . 0) =v(x) ut (x. 2 c Finally. 1 1 f (x) = (v(x) + W (x)). 1 u(x. 0) = cf (x) − cg (x) = w(x) which implies that 1 1 f (x) − g(x) = W (x) = c c Therefore. 0) =w(x) Solution.

110SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS Practice Problems Exercise 13.1 Show that if v(x.6 Solve the initial value problem utt =4uxx u(x. t) = ax + b is a solution to equation (13. t) and w(x.1).3 Find a solution to (13. 0) =1 ut (x. t) = u(L. 0) =0 Exercise 13.1) that satisﬁes the homogeneous conditions u(x. 0) = u(0.5 Solve the initial value problem utt =uxx 1 u(x. t) satisfy equation (13. 0) = cos x ut (x.2 Show that any linear time independent function u(x. where α and β are constants. 0) = 1 + x2 ut (x. t) = 0. 0) =0 Exercise 13.1). 0) = cos (2πx) .4 Solve the initial value problem utt =9uxx u(x.1) then αv + βw is also a solution to (13. Exercise 13. Exercise 13. Exercise 13.

0) =e−x ut (x. If u(x. u(x − y. (iii) Any dilation. is a solution. 0) = cos2 x 2 111 1 if x < 0 0 if x ≥ 0 Exercise 13. 0) =0 where v(x) = Exercise 13. 0) =v(x) ut (x. t).9 Prove that the wave equation.13 HYPERBOLIC TYPE: THE WAVE EQUATION Exercise 13. say ux (x. t) where y is a ﬁxed constant. utt = c2 uxx satisﬁes the following properties. t) = v(r) r cos nt is a solution to the PDE 2 urr + ur = utt . then (i) Any translate. r . (ii) Any derivative. at). u(ax. which are known as invariance properties.8 Solve the initial value problem utt =c2 uxx u(x. is also a solution. is also a solution. for any ﬁxed constant a. t) is a solution.7 Solve the initial value problem utt =25uxx u(x.10 Find v(r) if u(r. Exercise 13.

. ut (x. d > 0. 0) = sin x. t)).14 (a) Verify that for any R(x) the function u(x. show that the energy is constant. Exercise 13.12) are equal for the traveling wave solution in (a). dt (b) Assuming ﬁxed ends boundary conditions. for all t > 0.13 For a wave equation with damping utt − c2 uxx + dut = 0. t) − ut (0.12 The total energy of the string (the sum of the kinetic and potential energies) is deﬁned as 1 L 2 E(t) = (ut + c2 u2 )dx. t)ux (0. 0) = ex . Exercise 13.11 Find the solution of the wave equation on the real line (−∞ < x < +∞) with the initial conditions u(x. that is the ends of the string are ﬁxed so that u(0. t) = R(x − ct) is a solution of the wave equation utt = c2 uxx . Such solutions are called traveling waves. t)ux (L. t 2 0 (a) Using the wave equation derive the equation of conservation of energy dE(t) = c2 (ut (L. Exercise 13. (c) Assuming free ends boundary conditions for both x = 0 and x = L show that the energy is constant. t) = 0.112SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS Sample Exam Questions Exercise 13. t) = u(L. (b) Show that the potential and kinetic energies (see Exercise 13. 0 < x < L with the ﬁxed ends boundary conditions show that the total energy decreases.

ut (x.13 HYPERBOLIC TYPE: THE WAVE EQUATION Exercise 13. 0) = sin 2x.15 Find the solution of the Cauchy wave equation utt = 4uxx u(x. Simplify your answer as much as possible. 113 . 0) = x2 .

2. Before we begin our discussion of the mathematics of the heat equation. The most common units of heat are BTU (British Thermal Unit). and the only way heat can enter or leave the rod is at either end. we must ﬁrst determine what is meant by the term heat? Heat is type of energy known as thermal energy. placed along the x−axis from x = 0 to x = L as shown in Figure 14. Heat travels in waves like other forms of energy. an explosion releases light. Calorie and Joule. a burning candle releases light and heat waves.114SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS 14 Parabolic Type: The Heat Equation in OneDimensional Space In this section. Figure 14. We will look at a model for describing the distribution of temperature in a solid material as a function of time and space. It is often released along with other kinds of energy such as light waves or sound waves. On the other hand. .1 Assume the heat ﬂows only in the x−direction. Heat can be released through a chemical reaction (such as the nuclear reactions that make the Sun “burn”) or can be trapped for a limited time by insulators. with the lateral sides insulated. heat. Consider now a rod made of homogeneous heat conducting material of uniform density ρ and constant cross section A. For example. In other words.1. We will also assume that heat energy in any piece of the rod is conserved. Also we assume that the temperature of the rod is constant at any point of the cross section. and can change the matter it touches. and sound waves. Consider an inﬁnitesimal portion U of the rod from x to x + ∆x of length ∆x as shown in Figure 14. t) be the temperature of the cross section at the point x and the time t. It can heat it up and cause chemical reactions like burning to occur. temperature will only vary in x and we can hence consider the rod to be a one spatial dimensional rod. Let u(x.

14 PARABOLIC TYPE: THE HEAT EQUATION IN ONE-DIMENSIONAL SPACE115

Figure 14.2 From the theory of heat conduction, the rate of change of heat in the portion U is given by
x+∆x

cρAut (s, t)dx
x

where c is the speciﬁc heat, that is, the amount of heat energy that it takes to raise one unit of mass of the material by one unit of temperature. Assuming that u is continuously diﬀerentiable, we can apply the mean value theorem for integrals and ﬁnd x ≤ ξ ≤ x + ∆x such that
x+∆x

ut (s, t)dx = ∆xut (ξ, t).
x

Thus, the rate of change of heat in U is given by cρA∆xut (ξ, t). On the other hand, by Fourier (or Fick’s) law of heat conduction, the rate of heat ﬂow through any cross section is proportional to the area A and the gradient of the temperature normal to the cross section, and heat ﬂows in the direction of decreasing temperature. Thus, the rate of heat ﬂowing in U through the cross section at x is −KAux (x, t) and the rate of heat ﬂowing out of U through the cross section at x + ∆x is −KAux (x + ∆x, t), where K is the thermal conductivity of the rod. Now, the conservation of energy law states rate of change of heat in U = rate of heat ﬂowing in − rate of heat ﬂowing out

116SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS or mathematically written as, cρA∆xut (ξ, t) = −KAux (x, t) + KAux (x + ∆x, t) or cρA∆xut (ξ, t) = KA[ux (x + ∆x, t) − ux (x, t)]. Dividing this last equation by cAρ∆x and letting ∆x → 0 we obtain ut (x, t) = kuxx (x, t) (14.1)

K where k = cρ is called the diﬀusivity constant. Equation (14.1) is the one dimensional heat equation which is second order, linear, homogeneous, and of parabolic type. The non-homogeneous heat equation

ut = kuxx + f (x) is known as the heat equation with an external heat source f (x). An example of an exterenal heat source is the heat generated from a candle placed under the bar. The function
L

E(t) =
0

cρu(x, t)dx

is called the total thermal energy at time t of the entire rod. Example 14.1 The two ends of a uniform rod of length L are insulated. There is a constant source of thermal energy q0 = 0 and the temperature is initially u(x, 0) = f (x). (a) Write the equation and the boundary conditions for this model. (b) Calculate the total thermal energy of the entire rod. Solution. (a) The model is given by the PDE cρut (x, t) = Kuxx + q0 with boundary conditions ux (0, t) = ux (L, t) = 0.

14 PARABOLIC TYPE: THE HEAT EQUATION IN ONE-DIMENSIONAL SPACE117 (b) First note that d dt
L L L L

cρu(x, t)dx =
0 0

cρut (x, t)dx =
0

Kuxx dx +
0

q0 dx

= Kux |L + q0 L = 0 0 since ux (0, t) = ux (L, t) = 0. Integrating in time from 0 to t we ﬁnd E(t) = q0 Lt + C. But C = E(0) = energy is given by
L 0

cρu(x, 0)dx =
L

L 0

cρf (x)dx. Hence, the total thermal

E(t) =
0

cρf (x)dx + q0 Lt

Initial Boundary Value Problems In order to solve the heat equation we must give the problem some initial conditions. If you recall from the theory of ODE, the number of conditions required for solving initial value problems always matched the highest order of the derivative in the equation. In partial diﬀerential equations the same idea holds except now we have to pay attention to the variable we are diﬀerentiating with respect to as well. So, for the heat equation we have got a ﬁrst order time derivative and so we will need one initial condition and a second order spatial derivative and so we will need two boundary conditions. For the initial condition, we deﬁne the temperature of every point along the rod at time t = 0 by u(x, 0) = f (x) where f is a given (prescribed) function of x. This function is known as the initial temperature distribution. The boundary conditions will tell us something about what the temperature is doing at the ends of the bar. The conditions are given by u(0, t) = T0 and u(L, t) = TL . and they are called as the Dirichlet conditions. In this case, the general form of the heat equation initial boundary value problem is to ﬁnd u(x, t) satisfying ut (x, t) =kuxx (x, t), 0 ≤ x ≤ L, t > 0 u(x, 0) =f (x), 0 ≤ x ≤ L u(0, t) =T0 , u(L, t) = TL , t > 0.

118SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS In the case of insulated endpoints, i.e. there is no heat ﬂow out of them, we use the boundary conditions ux (0, t) = ux (L, t) = 0. These conditions are examples of what is known as Neumann boundary conditions. In this case, the general form of the heat equation initial boundary value problem is to ﬁnd u(x, t) satisfying ut (x, t) =kuxx (x, t), 0 ≤ x ≤ L, t > 0 u(x, 0) =f (x), 0 ≤ x ≤ L ux (0, t) =ux (L, t) = 0, t > 0.

14 PARABOLIC TYPE: THE HEAT EQUATION IN ONE-DIMENSIONAL SPACE119

Practice Problems
Exercise 14.1 Show that if u(x, t) and u(x, t) satisfy equation (14.1) then αu + βv is also a solution to (14.1), where α and β are constants. Exercise 14.2 Show that any linear time independent function u(x, t) = ax + b is a solution to equation (14.1). Exercise 14.3 Find a lnear time independent solution u to (14.1) that satisﬁes u(0, t) = T0 and u(L, T ) = TL . Exercise 14.4 Show that to solve (14.1) with the boundary conditions u(0, t) = T0 and u(L, t) = TL it suﬃces to solve (14.1) with the homogeneous boundary conditions u(0, t) = u(L, t) = 0. Exercise 14.5 Find a solution to (14.1) that satisﬁes the conditions u(x, 0) = u(0, t) = u(L, t) = 0. Exercise 14.6 Let (I) denote equation (14.1) together with intial time condition u(x, 0) = f (x), where f is not the zero function, and the homogeneous boundary conditions u(0, t) = u(L, t) = 0. Suppose a nontrivial solution to (I) can be written in the form u(x, t) = X(x)T (t). Show that X and T satisﬁes the ODE X − λ X = 0 and T − λT = 0. k for some constant λ. Exercise 14.7 Consider again the solution u(x, t) = X(x)T (t). Clearly, T (t) = T (0)e−λt . Suppose that λ > 0. √ √ (a) Show that X(x) = Ae αx + Be− αx , where α = λ and A and B are k arbitrary constants. √ (b) Show that A and B satisﬁes the two equations A + B = 0 and A(e λL − √ e− λL ) = 0.

L Exercise 14.11 Let ut = uxx for 0 < x < π and t > 0 with boundary conditions u(0. Let E(t) = 0 (u2 + u2 )dx. ux (L. t) = π 0 = u(π. √ √ (d) Using (b) and (c) show that e λL = e− λL .12 Suppose ut = uxx + 4. Exercise 14. k nπ L x sat- Exercise 14. t) = 5.10 Suppose that a wire is stretched between 0 and a.1) and the homogeneous boundary conditions.8 Consider the results of the previous exercise. t) = 6. t) and initial condition u(x.9 kn2 π 2 Show that u(x. 0) = f (x). . t) when (i) the left end is kept at 0 degrees and the right end is kept at 100 degrees. ux (0.120SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS (c) Show that A = 0 leads to a contradiction. Show that this equality leads to a contradiction. where un (x. 2 2 −λ . Exercise 14. We conclude that λ < 0. t) = cn e L2 t sin k=1 isﬁes (14. 0) = f (x). (a) Show that X(x) = c1 cos βx + c2 sin βx where β = (b) Show that λ = λn = − kn 2π . and (ii) when both ends are insulated. t) = n uk (x. where n is an integer. u(x. t x Show that E (t) < 0. Describe the boundary conditions for the temperature u(x. Exercise 14. t). Calculate the total thermal energy of the one-dimensional rod (as a function of time).

A function v(x) that satisﬁes the equation v (x) = 0. . the steady-state solutions of the heat equation are those solutions that don’t depend on time. 0) = f (x).13 Consider the heat equation ut = kuxx for x ∈ (0. u(L) = T. u(L) = T. (d) q(x) = 1. 1) and t > 0. k2 = 1. and satisﬁes u(0) = T1 . Exercise 14. ux (L) = α. u(0) = T1 . does not depend on time).14 PARABOLIC TYPE: THE HEAT EQUATION IN ONE-DIMENSIONAL SPACE121 Sample Exam Questions Exercise 14. Exercise 14. 0) = x for x ∈ (0. Find u(x). Find v(x). T2 = 100. Find the equilibrium (steady state) temperature distribution in the following cases. k1 = 2.15 Two rods of diﬀerent materials (thermal diﬀusion constants k1 and k2 . (a) q(x) = 0. with conditions v(0) = 2 and v(1) = 3 is called a steady-state solution. u(0) = T. t) = 3 for t > 0 and initial conditions u(x. Assume that the initial temperature distribution is given by u(x. 1). (b) Let L = 1. That is. (b) q(x) = 0.14 Consider the equation for the one-dimensional rod of length L with given heat energy source: ut = uxx + q(x). Plot the temperature u as a function of x. u(0) = 0. with boundary conditions u(0.) (a) Assume that the temperature distribution is steady (u = u(x). assume cρ = 1 in both cases) are joined at x = L and are in perfect thermal contact (this means that the temperature is continuous at x = L and the heat energy ﬂowing out of one end ﬂows into the other). u(2L) = T2 . (c) q(x) = 0. T1 = 0. ux (0) = 0. t) = 2 and u(1. (See ﬁgure below. u(L) = T2 .

0) = f (x). ux (L. (a) Give the expression for the total thermal energy of the rod. t) = 0. ux (0. Exercise 14. t) = ux (L.122SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS Exercise 14. t) = 7.16 Consider the equation for the one-dimensional rod of length L with insulated ends: cρut = Kuxx . ux (0. (a) Calculate the total thermal energy of the one-dimensional rod (as a function of time). u(x. (b) From part (a) ﬁnd the value of β for which a steady-state solution exist. t) = β. (c) For the above value of β ﬁnd the steady state solution. .17 Suppose ut = uxx + x. (b) Show using the equation and the boundary conditions that the total thermal energy is constant.

The expression on the right is called a trigonometric series. and the series provide a good approximation only in some (often small) vicinity of a reference point. etc. we say that a series of functions ∞ fn (x) converges uniformly n=1 to a function f if and only if the sequence of partial sums {Sn }∞ converges n=1 uniformly to f. there exist many types of convergence of series of functions. In this section we introduce a type of series of functions known as Fourier series. Fourier series constructed of trigonometric rather than power functions. and can be used for functions not only not diﬀerentiable. These functions must have inﬁnitely many derivatives. n=1 n→∞ Likewise. namely. The main limitation of Fourier series is that the underlying function should be periodic. Fourier series.1) where an and bn are called the Fourier coeﬃcients. They are given by nπ nπ a0 + an cos x + bn sin x f (x) = 2 L L n=1 ∞ . Recall from calculus that a function series is a series where the summands are functions. −L≤x≤L (15. Motivation: In Calculus we have seen that certain functions may be represented as power series by means of the Taylor expansions. We write ∞ fn (x) = lim Sn (x) = f (x). We say that a series of functions ∞ n=1 fn (x) converges pointwise to a function f if and only if the sequence of partial sums Sn (x) = f1 (x) + f2 (x) + · · · + fn (x) converges pointwise to f. etc. pointwise. Unlike series of numbers. The point here is to do just enough to allow us to do some basic solutions to partial diﬀerential equations later in the book. but even discontinuous at some points. uniform. Examples of function series include power series.15 AN INTRODUCTION TO FOURIER SERIES 123 15 An Introduction to Fourier Series In this and the next section we will have a brief look to the subject of Fourier series. Laurent series. Note that we begin the series with a0 as opposed to simply a0 to simplify the coeﬃcient formula for an that we 2 .

we look at some of the properties of Fourier series. Let f and g be two functions with domain the closed interval [a. Before answering these questions. We deﬁne . We want to deﬁne a similar concept for functions.1. Periodicity Property Recall that a function f is said to be periodic with period T > 0 if f (x + T ) = f (x) for all x.1) is periodic with fundamental period 2L. A graph of a T −periodic function is shown in Figure 15. We say that u and v are orthogonal if and only if u · v = 0. It is an easy exercise to show that the Fourier series (15.1 below. The main questions we want to consider next are the questions of determining which functions can be represented by Fourier series and if so how to compute the coeﬃcients an and bn . x + T in the domain of f. The smallest value of T for which f is periodic is called the fundamental period. Orthogonality Property Recall from Calculus that for each pair of vectors u and v we associate a scalar quantity u · v called the dot product of u and v. Note that the deﬁnite integral of a T −periodic function is the same over any interval of length T. Figure 15.1 For a T −periodic function we have f (x) = f (x + T ) = f (x + 2T ) = · · · .124SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS will derive later in this section. b]. if f and g are two periodic functions with common period T. By Exercise 15. then the product f g and an arbitrary linear combination c1 f + c2 g are also periodic with period T.

for n = m we have L cos −L mπ nπ 1 L (m + n)π x cos x dx = cos x + cos L L 2 −L L (m + n)π 1 L sin x = 2 (m + n)π L L + sin (m − n)π (m − n)π x L L (m − n)π x L dx =0 −L where we used the trigonometric identity 1 cos a cos b = [cos (a + b) + cos (a − b)]. Example 15. Symbolically. We call < f. −L Now. sin : n ∈ N is mutually orthogonal in L 1 · cos −L nπ nπ x dx = sin x L L nπ nπ x dx = − cos x L L L =0 −L L and L 1 · sin −L = 0. Solution. We say that f and g are orthogonal if and only if < f. we show that L sin −L mπ nπ x sin x dx = 0 L L .1 Show that the set 1. We have nπ x L nπ x L .15 AN INTRODUCTION TO FOURIER SERIES 125 a function that takes a pair of functions to a scalar. cos [−L. we write b < f. A set of functions is said to be mutually orthogonal if each distinct pair of functions in the set is orthogonal. L]. g > the inner product of f and g. g >= a f (x)g(x)dx. 2 In the exercises below. g >= 0.

L) we have a0 nπ nπ f (x) = + an cos x + bn sin x 2 L L n=1 ∞ ∞ L .1 (1) Almost all functions occurring in practice are piecewise smooth functions. b] execept possibly at ﬁnitely many points of discontinuity within the interval [a.126SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS and mπ nπ x sin x dx = 0 L L −L The reason we care about these functions being orthogonal is because we will exploit this fact to develop a formula for the coeﬃcients in our Fourier series. . cos Now. which functions can be expressed as a Fourier series expansion. ensures that a Fourier decomposition can be found for any function which is piecewise smooth. (2) Given a non-periodic function f on [−L. b] if it is continuous in [a. An example of a piecewise continuous function is the function x 0≤x<1 f (x) = x2 − x 1 ≤ x ≤ 2 We will say that f is piecewise smooth in [a. L]. L) we have a0 f (x− ) + f (x+ ) nπ nπ an cos = + x + bn sin x 2 2 L L n=1 where as for points of continuity x ∈ (−L. If f is a piecewise smooth on [−L. The following theorem. The above theorem applies to the periodic extension F of f where F (x + 2nL) = f (x) (n ∈ Z) and F (x) = f (x) on [−L. L] then for all points of discontinuity x ∈ (−L. b] if and only if f (x) as well as its derivatives are piecewise continuous. A function f (x) is said to be piecewise continuous on [a. that is. and at each point of discontinuity. the right. . L]. proven in more advanced books.1 Let f be a 2L-periodic function. Remark 15. Theorem 15.and lefthanded limits of f exist. b]. in order to answer the ﬁrst question mentioned earlier. we need to introduce some mathematical concepts.

n=1 (4) If f (x) is periodic. from the previous section. L]. we will answer the second question mentioned earlier. that is.15 AN INTRODUCTION TO FOURIER SERIES 127 Convergence Results of Fourier Series We list few of the results regarding the convergence of Fourier series: (1) The type of convergence in the above theorem is pointwise convergence. −L Thus. Euler-Fourier Formulas Next. L] such that f (−L) = f (L). Integrating both sides of (15. f (x)dx.1) converges uniformly to f (x) on the interval [−L.1) we obtain L L f (x)dx = −L −L a0 dx + 2 L ∞ an cos −L n=1 nπ nπ x + bn sin x L L dx. These formulas for an and bn are called Euler-Fourier formulas which we derive next. But L cos −L nπ nπ L x dx = sin x L nπ L nπ L nπ x dx = − cos x L nπ L 1 a0 = L L =0 −L and likewise L L sin −L = 0. (5) The convergence is uniform on any closed interval that does not contain a point of discontinuity. then the Fourier Series corresponding to f converges uniformly to f (x) for the entire real line. and has a piecewise continuous derivative. the question of ﬁnding formulas for the coeﬃcients an and bn . continuous. Since the trigonometric series is assumed to be uniformly convergent. (3) The convergence is uniform whenever ∞ (|an |2 + |bn |2 ) is convergent. (2) The convergence is uniform for a continuous function f on [−L. −L . We will assume that the RHS in (15. we can interchange the order of integration and summation to obtain L L f (x)dx = −L −L a0 dx + 2 n=1 ∞ L an cos −L nπ nπ x + bn sin x L L L dx.

n.128SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS To ﬁnd the other Fourier coeﬃcients. we multiply both sides of (15. ∀m. x > 0. we can show that bm = 1 L L f (x) cos −L L f (x) sin −L mπ x dx.2 . we recall the results of Exercises 15. to ﬁnd the formula for the Fourier coeﬃcients am for m > 0. x ≤ 0 x.1) by cos mπ x and integrate from −L to L to otbain L mπ x = f (x) cos L −L L L −L a0 mπ cos x dx + an 2 L n=1 L ∞ L cos −L nπ mπ x cos x dx L L + bn −L sin nπ mπ x cos x L L mπ x dx = am L L mπ x dx. L L Now.2 Find the Fourier series expansion of f (x) = on the interval [−π. Hence. L Example 15.3 below. L cos −L L nπ mπ x cos x dx = L L nπ mπ x sin x dx = L L sin L if m = n 0 if m = n L if m = n 0 if m = n sin −L L −L nπ mπ x sin x dx = 0. L f (x) cos −L and therefore 1 am = L Likewise.15. π]. 0. . L dx.

Let F be a periodic extension of f of period 2π.3 Apply the Theorem 15. π (−1)n − 1 (−1)n+1 f (x) = + cos (nx) + sin (nx) 4 n=1 πn2 n Example 15.2. F is a piecewise smooth function so that by the previous thereom we can write  π ∞ if x = −π  2. n n+1 (−1) − 1 (−1) π f (x). π]. f (x) = F (x) on the interval [−π.15 AN INTRODUCTION TO FOURIER SERIES Solution. (2n − 1)2 8 This provides a method for computing an approximate value of π . We have 1 a0 = π 1 an = π 1 bn = π Hence. ∞ 129 1 f (x)dx = π −π π π π 0 xdx = 1 π 0 π 1 x sin nxdx = π 0 x cos nxdx = π 2 π x sin nx cos nx (−1)n − 1 + = n n2 0 πn2 π x cos nx sin nx (−1)n+1 − + = n n2 0 n Solution. if −π < x < π + cos (nx) + sin (nx) =  π 4 n=1 πn2 n . Clearly. if x = π 2 Taking x = π we have the identity π (−1)n − 1 π + (−1)n = 2 4 n=1 πn 2 which can be simpliﬁed to ∞ ∞ n=1 1 π2 = .1 to the function in Example 15. Thus.

L]. Answers to these questions are provided next. Theorem 15.2 A Fourier series of a piecewise smooth function f can always be integrated term by term and the result is a convergent inﬁnite series that always conL verges to −L f (x)dx even if the original series has jumps.2 An example of a function that does not a Fourier series representation is 1 the function f (x) = x2 on [−L. . The ﬁnal topic of discussion here is the topic of diﬀerentiation and integration of Fourier series. The result of the diﬀerentiation is the Fourier series of f (x). Thus. Likewise we want to know if we can integrate a Fourier series term by term and arrive at the Fourier series of the integral of the function. Theorem 15. In particular we want to know if we can diﬀerentiate a Fourier series term by term and have the result be the Fourier series of the derivative of the function. the coeﬃcient a0 for this function does not exist.130SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS Remark 15.3 A Fourier series that is continuous can be diﬀerentiated term by term if f (x) is piecewise smooth. For example. not every function can be written as a Fourier series expansion.

x = 0.3 Compute the following integrals: L (a) −L cos2 nπ x dx.5 Find the Fourier series of f (x) = x2 − 1 2 on the interval [−1. .2 Show that for m = n we have (a) (b) L sin mπ x L −L L cos mπ x L −L sin sin nπ x L nπ x L dx = 0 and dx = 0 Exercise 15. π on the interval [−π.4 Find the Fourier coeﬃcients of   −π. π < x < 2π. (b) c1 f + c2 g is periodic of period T. 1]. 0<x<π f (x) =  0. Exercise 15. −π < x < π f (x) =  1. L L (b) −L sin2 nπ x dx. −2π < x < −π 0. L L nπ (c) −L cos L x sin nπ x dx. L Exercise 15. where c1 and c2 are real numbers. −π ≤ x < 0 π.1 Let f and g be two functions with common domain D and common period T.6 Find the Fourier series of the function   −1. Exercise 15. Exercise 15. Show that (a) f g is periodic of period T. π].15 AN INTRODUCTION TO FOURIER SERIES 131 Practice Problems Exercise 15.

1]. (ii) Discuss its limit: In particular.9 Assume that f (x) is continuous and has period 2L. 0 < x ≤ 2.10 Consider the function f (x) deﬁned by f (x) = 1 0≤x<1 2 1≤x<3 and extended periodically with period 3 to R so that f (x + 3) = f (x) for all x. 1 x is not piecewise continuous on [−1. it does not matter over which interval the Fourier coeﬃcients are computed as long as the interval length is 2L. In particular. (i) Find the Fourier series of f (x). −2 ≤ x ≤ 0 1 − x.132SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS Exercise 15.8 Show that f (x) = 1 + x. . and what is this limit? (iii) Plot the graph of f (x) and the limit of the Fourier series. Prove that L L+a f (x)dx = −L −L+a f (x)dx is independent of a ∈ R. [Remark: This result is also true for piecewise continuous functions]. Exercise 15. does the Fourier series converge pointwise or uniformly to its limit.7 Find the Fourier series of the function f (x) = Exercise 15. Exercise 15.

(b) f (x) = 2 + sin pix . 2. determine the coeﬃcients an . 1. To what value does the Fourier series of f (x) converge when x = 0? Exercise 15. L 1 x≤0 (c) f (x) = (d) f (x) = x. Find the Fourier coeﬃcients an .13 Let f (x) = x3 on [−π.14 Let f (x) be the square wave function f (x) = −π −π ≤ x < 0 π 0≤x≤π extended periodically to all of R. 0 x>0 Exercise 15. 2. 3.15 (a) Find the Fourier series of f (x) = 1 −π ≤ x < 0 2 0≤x≤π extended periodically to all of R. π]. n = 1. n ∈ N of the Fourier series expansion. · · · and bn .12 Let f (t) be the function with period 2π deﬁned as f (t) = 2 if 0 ≤ x ≤ π 2 0 if π < x ≤ 2π 2 f (t) has a Fourier series and that series is equal to a0 + ∞(an cos nt + bn sin nt). 2 . extended periodically to all of R. 2 n=1 Find a3 . b3 Exercise 15. Exercise 15. Simplify your coeﬃcients as much as possible. n+1 (b) Use (a) to evaluate the series ∞ (−1) . n = 0. Hint: Evaluate the Fourier n=1 (2n−1) series at x = π .11 For the following functions f (x) on the interval −L < x < L.15 AN INTRODUCTION TO FOURIER SERIES 133 Sample Exam Questions Exercise 15. · · · . (a) f (x) = 1.

That is. Then (f g)(−x) = f (−x)g(−x) = [−f (x)][−g(x)] = (f g)(x).134SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS 16 Fourier Sines Series and Fourier Cosines Series In this section we discuss some important properties of Fourier series when the underlying function f is either even or odd. Now. (a) Suppose that both f and g are even. Then (f g)(−x) = f (−x)g(−x) = −f (x)g(x) = −(f g)(x). (a) Since the function sin nπ x L L f (x) sin nπ x L −L L nπ f (x) cos L x −L dx when f is even. f g is even. f g is odd Example 16. That is. we have that f (x) sin . Solution. we recall from Exercises (1. A function f is odd if it satisﬁes f (−x) = f (x) for all x in the domain of f whereas f is even if it satisﬁes f (−x) = f (x) for all x in the domain of f.6)-(1. nπ x L is odd and f is even. Then (f g)(−x) = f (−x)g(−x) = f (x)g(x) = (f g)(x). suppose that both f and g are odd.2 (a) Find the value of the integral (b) Find the value of the integral Solution. That is. f g is even. dx when f is odd. If f (x) is even then L L f (x)dx = 2 −L 0 f (x)dx. (b) If f is odd and g is even then f g is odd. Example 16. (b) f is odd and g is even. Now.1 Show the following (a) If f and g are either both even or both odd then f g is even. −L Using just these basic facts we can ﬁgure out some important properties of the Fourier series we get for odd or even functions.7) the following facts about even and odd functions. If f is odd then L f (x)dx = 0.

We can then further extend this function to the entire real line by deﬁning it to be 2L periodic. x=0 This function will be odd on the interval [−L. L]. The coeﬃcients bn are given by the formula bn = 1 L 2 = L nπ 2 x dx = L L −L L nπ f (x) sin x dx L 0 f odd sin L L f odd sin 0 nπ x dx L since f odd sin nπ x is an even function. L]. Moreover. L Likewise. L] we have ∞ f (x) = n=1 bn sin nπ x . we have that f (x) cos nπ x dx = 0 L f (x) cos −L Even and Odd Extensions Let f : [0.16 FOURIER SINES SERIES AND FOURIER COSINES SERIES is odd so that 135 L f (x) sin −L nπ x dx = 0 L nπ x L (b) Since the function cos is odd so that nπ x L L is even and f is odd. in the interval [0. We note that f odd is an odd function and piecewise smooth so that by Theorem 15. We deﬁne the odd extension of this function on the interval −L ≤ x ≤ L by  0<x≤L  f (x) −f (−x) −L ≤ x < 0. we can deﬁne the even extension of f on the interval −L ≤ x ≤ L by f (x) 0≤x≤L feven (x) = f (−x) −L ≤ x ≤ 0 . and from the fact that it is odd all of the an s are zero. L (16. fodd (x) =  0. Let f odd denote this extension. and will be equal to f (x) on the interval [0.1 it possesses a Fourier series expansion.1) the Fourier sine series of f.1) We call (16. L] → R be a piecewise smooth function.

136SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS We can then further extend this function to the entire real line by deﬁning it to be 2L periodic. Let f even denote this extension. Again, we note that f even is equal to the original function f (x) on the interval upon which f (x) is deﬁned. Since f even is piecewise smooth, by Theorem 15.1 it possesses a Fourier series expansion, and from the fact that it is even all of the bn s are zero. Moreover, in the interval [0, L] we have f (x) = nπ a0 + an cos x . 2 L n=1

(16.2)

We call (16.2) the Fourier cosine series of f. The coeﬃcients an are given by 2 L nπ an = f (x) cos x dx. L 0 L Example 16.3 Graph the odd and even extensions of the function f (x) = x, 0 ≤ x ≤ 1. Solution. We have fodd (x) = x for −1 ≤ x ≤ 1. The odd extension of f is shown in Figure 18.1(a). Likewise, feven (x) = x 0≤x≤1 −x −1 ≤ x ≤ 0

The even extension is shown in Figure 16.1(b)

Figure 16.1

16 FOURIER SINES SERIES AND FOURIER COSINES SERIES Example 16.4 Find the Fourier sine series of the function f (x) = Solution. We have 2 bn = π
0

137

x, 0≤x≤ π 2 π − x, π ≤ x ≤ π 2

π 2

π

x sin nxdx +
π 2

(π − x) sin nxdx .

Using integration by parts we ﬁnd
π 2 π x 1 2 2 x sin nxdx = − cos nx + cos nxdx n n 0 0 π π cos (nπ/2) 1 2 =− + 2 [sin nx]0 2n n π cos (nπ/2) sin (nπ/2) + =− 2n n2 π

0

while
π
π 2

(π − x) (π − x) sin nxdx = − cos nx n =

π
π 2

1 − n

π

cos nxdx
π 2

π cos (nπ/2) 1 − 2 [sin nx]π π 2 2n n π cos (nπ/2) sin (nπ/2) = + 2n n2 Thus, 4 sin (nπ/2) , πn2 and the Fourier sine series of f (x) is bn =

f (x) =
n=1

4 sin (nπ/2) sin nx = πn2

n=1

4(−1)2n−1 sin (2n − 1)x π(2n − 1)2

138SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS

Practice Problems
Exercise 16.1 Give an example of a function that is both even and odd. Exercise 16.2 Graph the odd and even extensions of the function f (x) = 1, 0 ≤ x ≤ 1. Exercise 16.3 Graph the odd and even extensions of the function f (x) = L − x for 0 ≤ x ≤ L. Exercise 16.4 Graph the odd and even extensions of the function f (x) = 1 + x2 for 0 ≤ x ≤ L. Exercise 16.5 Find the Fourier cosine series of the function f (x) = x, 0≤x≤ π 2 π − x, π ≤ x ≤ π 2

Exercise 16.6 Find the Fourier cosine series of f (x) = x on the interval [0, π]. Exercise 16.7 Find the Fourier sine series of f (x) = 1 on the interval [0, π]. Exercise 16.8 Find the Fourier sine series of f (x) = cos x on the interval [0, π]. Exercise 16.9 Find the Fourier cosine series of f (x) = e2x on the interval [0, 1]. Exercise 16.10 Consider the function f (x) = x on the interval [0, 2]. (i) Write down the deﬁnition of the Fourier sine and cosine series for f, including formulas for the coeﬃcients in terms of f. (Note: you do not need to compute the coeﬃcients by carrying out the integration. Just write down their formulas in terms of f.)

16 FOURIER SINES SERIES AND FOURIER COSINES SERIES

139

(ii) What is the limit of the Fourier sine series of f on (0, 2) and in what sense does it converge to its limit? (iii) What is the limit of the Fourier cosine series of f on (0, 2) and in what sense does it converge to its limit? (iv) Plot the even extension of f (x), the odd extension of f (x), the limit of its Fourier sine series, and the limit of its Fourier cosine series on the interval [−2, 2]. (Don’t forget to consider the endpoints 2 and −2.)

140SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS

Sample Exam Questions
Exercise 16.11 For the following functions on the interval 0 < x < L, ﬁnd the coeﬃcients bn of the Fourier sine expansion. (a) f (x) = sin 2π x . L (b) f (x) = 1 π (c) f (x) = cos L x . Exercise 16.12 For the following functions on the interval −L < x < L, ﬁnd the coeﬃcients an of the Fourier cosine expansion. (a) f (x) = 5 + cos 5π x . L (b) f (x) = x (c) 1 0<x≤ L 2 f (x) = 0 x> L 2 Exercise 16.13 Consider a function f (x), deﬁned on 0 ≤ x ≤ L, which is even (symmetric) around x = L . Show that the even coeﬃcients (n even) of the Fourier sine 2 series are zero. Exercise 16.14 Consider a function f (x), deﬁned on 0 ≤ x ≤ L, which is odd around x = L . 2 Show that the even coeﬃcients (n even) of the Fourier cosine series are zero. Exercise 16.15 The Fourier sine series of f (x) = cos πx cos = L where b1 = 0, bn =

πx L

for 0 < x < L is given by nπx , n∈N L

bn sin
n=1

2n [1 + (−1)n ]. 2 − 1)π (n
nπx L

Using term-by-term integration, ﬁnd the Fourier cosine series of sin

.

(c) Find an (n = 1. (b) Find a0 in the Fourier series for the even extension of f. (e) Write the Fourier series for the even extension of f. · · · ) in the Fourier series for the even extension of f. (d) Find bn in the Fourier series for the even extension of f.16 Consider the function f (x) = 1 0≤x<1 2 1≤x<2 141 (a) Sketch the even extension of f. . 2.16 FOURIER SINES SERIES AND FOURIER COSINES SERIES Exercise 16.

one assumes that it is possible to separate the contributions of the independent variables into separate functions that each involve only one independent variable. and divide the resulting equation by X(x)Y (y). The general solution of the ODEs is found. In developing a solution to a partial diﬀerential equation by separation of variables. ut = X(x)T (t) and uxx = X (x)T (t) this is basically a consequence of the fact that diﬀerentiation with respect to x sees t as a constant. y) = X(x)Y (y). If you are lucky your PDE has a solution with separable variables. Most of the PDE techniques involve a mixture of analytic. Then the problem turns into a set of separated ODEs (one for X(x) and one for Y (y). Of course. qualitative and numeric approaches. u(x. Thus. t) = X(x)T (t) of the eqaution uxx − ux = ut Solution. u(x.1 Find all the solutions of the form u(x. Example 17. and boundary initial conditions are imposed. the method consists of the following steps 1. each of the factors being a function of one independent variable. 3. Substitute into the PDE. It is very easy to ﬁnd the derivatives of a separable function: ux = X (x)T (t). there are some easy PDEs too.142SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS 17 Separation of Variables for PDEs Finding analytic solutions to PDEs is essentially impossible. .) 4. Now the equation uxx − ux = ut becomes X (x)T (t) − X (x)T (t) = X(x)T (t). y) is formed by multiplying together X(x) and Y (y). That is. 2. Factorize the (unknown) dependent variable of the PDE into a product of functions. We illustrate these steps in the next two examples. 5. and vice versa. In this section we discuss the application of the method of separation of variables in the solution of PDEs.

2 −(1 + 4λ) x 2 + Be 2 +λt sin x −(1 + 4λ) x 2 .17 SEPARATION OF VARIABLES FOR PDES We can separate variables further. u(x. That is. Division by X(x)T (t) gives T (t) X (x) − X (x) = . The ﬁrst equation is solved via the characteristic equation ω 2 − ω − λ = 0. X(x) T (t) Thus. whose solutions are √ 1 ± 1 + 4λ ω= . t −(1 + 4λ) x 2 + Be 2 sin x −(1 + 4λ) x . u(x. X(x) T (t) 143 The expression on the LHS is a function of x whereas the one on the RHS is a function on t only. t) = De 2 +λt cos x x x x u(x. 2 If λ > − 1 then 4 X(x) = Ae In this case. They both have to be constant. we have the following ODEs: X − X − λX = 0 and T = λT. The second equation is easy to solve: T (t) = Ceλt . eλt . t) = De 1 If λ = − 4 then √ 1+ 1+4λ x 2 √ 1− 1+4λ x 2 + Be . X (x) − X (x) T (t) = = λ. t) = (D + Ex)e 2 − 4 . √ 1+ 1+4λ x 2 e + Ee x λt √ 1− 1+4λ x 2 X(x) = Ae 2 + Bxe 2 and in this case If λ < − 1 then 4 X(x) = Ae 2 cos In this case.

Assuming X(x)Y (y) is nonzero. y) = X(x)Y (y). The solutions of these equations depend on the sign of λ. • If λ > 0 then the solutions are given X(x) =Ae + Be− λx √ √ Y (y) =C cos λy + D sin λy λx √ √ Y (y) Y (y) where A. we ﬁnd: X (x) Y (y) =− .144SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS Example 17. t) =k1 e λx cos λy + k2 e λx sin λy √ √ √ √ +k3 e− λx cos λy + k4 e− λx sin λy.2 Solve Laplace’s equation using the separation of variables method ∆u = uxx + uyy = 0. dividing for X(x)Y (y) and subtracting from both sides. Substituting in the Laplace’s equation. and D are constants. we obtain X (x)Y (y) + X(x)Y (y) = 0. B. • If λ = 0 then X(x) =Ax + B Y (y) =Cy + D . C. We look for a solution of the form u(x. In this case. That is. Solution. This says that they must equal to a constant. √ √ √ √ u(x. This results in the following two ODEs X − λX = 0 and Y + λY = 0. Y (y) X (x) =− =λ X(x) Y (y) where λ is a constant. X(x) Y (y) The left hand side is a function of x while the right hand side is a function of y.

B. and D are arbitrary constants. C. In this case. In this case.17 SEPARATION OF VARIABLES FOR PDES where A. Substitute u(x. xX yY The left hand side is a function of x while the right hand side is a function of y. This can be separated into Y X = . • If λ < 0 then √ √ X(x) =A cos −λx + B sin −λx Y (y) =Ce √ −λy √ 145 + De− −λy where A. t) = X(x)Y (y) into the given equation we ﬁnd yX Y − xXY = 0. y) = k1 xy + k2 x + k3 y + k4 . √ √ √ √ u(x. That is. This says that they must equal to a constant. xX yY where λ is a constant. This results in the following two ODEs X − λxX = 0 and Y − λyY = 0. yux − xuy = 0. B. the general solution is given by u(x. u(x. Solution. Thus. Solving these equations using the method of separation of variable for ODEs λy 2 λx2 we ﬁnd X(x) = Ae 2 and Y (y) = Be 2 . X Y = = λ. and C are arbitrary constants. y) = Ce λ(x2 +y 2 ) 2 .3 Solve using the separation of variables method. y) =k1 cos −λxe −λy + k2 cos −λxe− −λy √ √ √ √ +k3 sin −λxe −λy + k4 sin −λxe− −λy Example 17.

0 < x < 1. . Exercise 17. t) = X(x)T (t). t) t > 0 of the form u(x. (Note: you do not need to solve for X and T . u(0. r r Exercise 17.2 Solve using the separation of variables method ut = kuxx .1 Solve using the separation of variables method ∆u + λu = 0 Exercise 17. t) t > 0 of the form u(x. 0) = f (x). t > 0 u(0. t) = 0 = ux (1.3 Derive the system of ordinary diﬀerential equations for X(x) and T (t) that is satisﬁed by solutions to 1 1 urr + ur + 2 uθθ = 0. (Note: you do not need to solve for X and T .) Exercise 17. 0 < x < L. t > 0 u(x.) Exercise 17. t) = X(x)T (t). t) = 0 = ux (L.5 Derive the system of ordinary diﬀerential equations and boundary conditions for X(x) and T (t) that is satisﬁed by solutions to ut = kuxx .6 Find all product solutions of the PDE ux + ut = 0.4 Derive the system of ordinary diﬀerential equations and boundary conditions for X(x) and T (t) that is satisﬁed by solutions to utt = uxx − 2u.146SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS Practice Problems Exercise 17.

17 SEPARATION OF VARIABLES FOR PDES 147 Exercise 17.7 Derive the system of ordinary diﬀerential equations for X(x) and Y (y) that is satisﬁed by solutions to 3uyy − 5uxxxy + 7uxxy = 0. y) = X(x)Y (y). of the form u(x. uxy + u = 0. Exercise 17.8 Find the general solution by the method of separation of variables. Exercise 17. ux − yuy = 0. .9 Find the general solution by the method of separation of variables.

Exercise 17.13 Separate the PDE uxx − uy + uyy = u into two ODEs with a parameter.11 For the following PDEs ﬁnd the ODEs implied by the method of separation of variables. ux − uy = 0. utt − uxx = 0. Exercise 17. Exercise 17.12 Find all solutions to the following partial diﬀerential equation that can be obtained via the separation of variables. .148SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS Sample Exam Questions Exercise 17.10 Find the general solution by the method of separation of variables. You do not need to solve the ODEs. (a) ut = kr(rur )r (b) ut = kuxx − αu (c) ut = kuxx − aux (d) uxx + uyy = 0 (e) ut = kuxxxx .

Let’s assume that the solution can be written in the form u(x. . t) = u(L. t) = 0 = X(0)T (t) =⇒ X(0) = 0 and u(L. t) = X(x)T (t). t) = T0 and u(L. 0) = f (x) and the Dirichlet boundary conditions u(0. This gives the two ordinary diﬀerential equations X − λX = 0 and T − kλT = 0. X kT Since the LHS only depends on x and the RHS only depends on t. we have u(0.4. then there must be a constant λ such that X X = λ and T kT = λ. we consider the three cases of the sign of λ. As far as the boundary conditions. t) = 0 = X(L)T (t) =⇒ X(L) = 0. t) = TL .18 SOLUTIONS OF THE HEAT EQUATION BY THE SEPARATION OF VARIABLES METHOD149 18 Solutions of the Heat Equation by the Separation of Variables Method In this section we apply the method of separation of variables in solving the one spatial dimension of the heat equation. Substituting into the heat equation we obtain T X = . it suﬃces to solve the problem with the Dirichlet boundary conditions being replaced by the homogeneous boundary conditions u(0. The Heat Equation with Dirichlet Boundary Conditions Consider the problem of ﬁnding all nontrivial solutions to the heat equation ut = kuxx that satisﬁes the initial time condition u(x. Next. t) = 0 (that is. the endpoints are assumed to be at zero temperature) with u not the trivial solution. Note that T is not the zero function for otherwise u ≡ 0 and this contradicts our assumption that u is the non-trivial solution. From Exercise 14.

Since X(0) = 0 we ﬁnd b = 0. Thus. n ∈ N.150SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS Case 1: λ = 0 In this case. we obtain sin −λL = 0 or −λL = nπ where n ∈ N. We must have B = 0 otherwise√ X(x) = 0 and this leads to the trivial solution. the functions n2 π 2 nπ x e− L2 kt .1) to obtain ∞ f (x) = n=1 Cn sin nπ x . X = 0. Since X(L) = 0 we ﬁnd a = 0. t) = u(L. u is the trivial solution. Case 2: λ > 0 √ √ In this case. L (18. n ∈ N. The condition X(L) = 0 implies B sin −λL = 0. we invoke the superposition principle of linear PDE to write n2 π 2 kt L2 nπ x. The√ condition X(0) = 0 implies A = 0. L . X ≡ 0 and u(x. X(x) = Ae λx + Be− λx . 0) = f (x). in order for these solutions to satisfy the initial value condition u(x. Case 3: λ < 0 √ √ In this case. we obtain inﬁnitely many solutions given by 2 Xn (x) = An sin Now. L .1) To determine the unknown constants Cn we use the initial condition u(x. t) = Cn sin ∞ u(x. Solving for λ 2 2 we ﬁnd λ = − nLπ . Since √ B = 0. solving the equation T − λkT = 0 by the method of separation of variables we obtain Tn (t) = Bn e− Hence. Solving this equation we ﬁnd X(x) = ax + b. t) = n=1 Cn sin n2 π 2 nπ x e− L2 kt . Now. Again. X(x) = A cos −λx + B sin −λx. Hence. t) = 0. the conditions X(0) = X(L) = 0 imply A = B = 0 and hence the solution is the trivial solution. t) ≡ 0. 0) = f (x) in (18. n ∈ N L satisfy ut = kuxx and the boundary conditions u(0. un (x. That is.

X (L)T (t) implis X (L) = 0. n = 0. the solution to the heat equation with nonhomogeneous condition u(0. √ √ √ The conditions X (0) = X (L) = 0 imply −λB = 0 and −λA sin −λL = 2 2 0.2). the coeﬃcients Cn are given by Cn = 2 L L f (x) sin 0 nπ x dx. t) = TL is given by ∞ u(x. the solution to the heat equation is given by (18. Hence. · · · L . 0 ≤ x ≤ L.1 According to Exercise 14. L]. 0 ≤ x ≤ L ux (0. diﬀerentiating X(x) = A cos −λx + √ B sin −λx with respect to x we ﬁnd √ √ √ √ X (x) = − −λA sin −λx + −λB cos −λx. In this case. Remark 18. Now. t) = n=1 Cn sin n2 π 2 nπ TL − T0 x e− L2 kt + T0 + x. t) = 0.1) with the Cn s calculated from (18. B = 0 and λ = − nLπ and 2 Xn (x) = An cos nπ x . Since 0 = ux (0.4. Likewise. t) = ux (L. 0) =f (x). t) =ux (L. 1. we use the boundary conditions ux (0. t > 0. t) = X (0)T (t) we obtain X (0) = 0. 2. 0 = u√ t) = x (L. t) satisfying ut (x. that is. t) = T0 and u(L. t > 0 u(x. L (18. t) = 0. there is no heat ﬂow out of them.2) Thus. t). L L The Heat Equation with Neumann Boundary Conditions When both ends of the bar are insulated.18 SOLUTIONS OF THE HEAT EQUATION BY THE SEPARATION OF VARIABLES METHOD151 Since the right-hand side is the Fourier sine series of f on the interval [0. t) =kuxx (x. the general form of the heat equation initial boundary value problem is to ﬁnd u(x.

L . t) = n=0 Cn cos n2 π 2 nπ x e− L2 kt L where 1 C0 = L 2 Cn = L L f (x)dx 0 L f (x) cos 0 nπ x dx.152SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS and n2 π 2 nπ x e− L2 kt . L By the superposition principle. n ∈ N. t) = Cn cos ∞ u(x. the required solution to the heat equation with Neumann boundary conditions is given by un (x.

2 Exercise 18. Exercise 18.1 Find the temperature in a bar of length 2 whose ends are kept at zero and lateral surface insulated if the initial temperature is f (x) = sin π x + 2 3 sin 5π x .4 Solve the following heat equation with Dirichlet boundary conditions ut = kuxx u(0. t) = ux (L. t) = 0 u(x.2 Find the temperature in a homogeneous bar of heat conducting material of length L with its end points kept at zero and initial temperature distribution dx given by f (x) = L2 (L − x). t) = 0 u(x.6 Solve ut = kuxx subject to ux (0. 0) = Exercise 18.18 SOLUTIONS OF THE HEAT EQUATION BY THE SEPARATION OF VARIABLES METHOD153 Practice Problems Exercise 18. t) = 0 9π u(x. Exercise 18. 0) = 0 0<x< L 2 1 L ≤x<L 2 1 0<x< L 2 2 L ≤x<L 2 . t) = u(L.5 Solve ut = kuxx u(0.3 Find the temperature in a thin metal rod of length L. t) = u(L. L Exercise 18. with both ends insulated (so that there is no passage of heat through the ends) and with initial π temperature in the rod f (x) = sin L x . 0) = 6 sin x .

10 Find the general solution u(x. t). 0 < x < L. t) = 0 u(x.9 Find the general solution of the following equation and brieﬂy describe its limit as t → ∞. L 3π x . 0 < x < L. Exercise 18. . ut = uxx . t) = 0 = ux (L. t > 0 ux (0. Brieﬂy describe its behavior as t → ∞. t) of ut = uxx − u. L Exercise 18. t) = 0 u(x. t) = 0 = u(L. 0) = −3 cos 8π x . t). 0) = 6 + 4 cos Exercise 18.8 Solve ut = kuxx subject to ux (0.7 Solve ut = kuxx subject to ux (0. t > 0. t) = ux (L. t ∈ R. t ∈ R ux (0.154SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS Exercise 18. t) = ux (L.

11 (Energy method) Let u1 and u2 be two solutions to the Robin boundary value problem ut = uxx − u. 0 < x < 1 Deﬁne w(x. derive the ODEs for X and T.18 SOLUTIONS OF THE HEAT EQUATION BY THE SEPARATION OF VARIABLES METHOD155 Sample Exam Questions Exercise 18. w(x. t) = u(L. (a) Show that w satisﬁes the initial value problem wt = wxx − w.12 Consider the heat induction in a bar where the left end temperature is maintained at 0. t) = 0. conclude that u1 = u2 . t > 0 u(x. 0 < x < 1 (b) Deﬁne E(t) = w2 (x. Can you choose a value of λ so that the other boundary condition is also satisﬁed? Exercise 18.13 Using the method of separation of variables ﬁnd the solution of the heat equation ut = kuxx satisfying the following boundary and initial conditions: (a) u(0. t). and the right end is perfectly insulated. t) = u(L. t) = 0. 0 ≤ E(t) ≤ E(0) for all t > 0. 0) = 0. 0 < x < 1. (a) Derive the boundary conditions of the temperature at the endpoints. u(x. (b) Following the separation of variables approach. u(x. t) = ux (1. 0) = 3 sin πx − sin 3πx L L 1 0 . What are the values of X(0) and X(1)? √ Show that solutions of the form X(x) = sin −λx satisﬁy the ODE and one of the boundary conditions. (c) Consider the equation in X(x). t) − u2 (x. (c) Show that E(t) = 0. Show that E (t) ≤ 0. t) = 0. t > 0 w(x. Exercise 18. 0) = g(x). Hence. t) = 0. 0) = 6 sin 9πx L (b) u(0. t) = u1 (x. 0 < x < 1. t > 0 ux (0. We assume k = 1 and L = 1. Hence. t)dx ≥ 0 for all t ≥ 0.

156SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS Exercise 18. t > 0 u(0. L L (b) ux (0. u(x. . t) = 0. Exercise 18. t) = ux (L. 0) = cos πx + 4 cos 5πx . 0 < x < 4π.15 Find the solution of the following heat conduction partial diﬀerential equation ut = 8uxx . u(x. 0 < x < 4π. t) = 0.14 Using the method of separation of variables ﬁnd the solution of the heat equation ut = kuxx satisfying the following boundary and initial conditions: (a) ux (0. 0) = 5. t) = ux (L. 0) = 6 sin x. t) = u(4π. t > 0 u(x. t) = 0.

Physically speaking. y) = (0. We have 2x + y2 2y 2 − 2x2 uxx = 2 (x + y 2 )2 2y uy = 2 x + y2 2x2 − 2y 2 uyy = 2 (x + y 2 )2 ux = x2 Plugging these expressions into the equation we ﬁnd uxx + uyy = 0. a boundary-value problem consists of ﬁnding a function which satisﬁes a given partial diﬀerential equation and particular boundary conditions. 0). for all (x. (19.19 ELLIPTIC TYPE: LAPLACE’S EQUATIONS IN RECTANGULAR DOMAINS157 19 Elliptic Type: Laplace’s Equations in Rectangular Domains Boundary value problems are of great importance in physical applications. Example 19. Solution. involving only space coordinates. u(x. y) = ln (x2 + y 2 ) is a harmonic function. The main model example of an elliptic type PDE is the Laplace equation ∆u = uxx + uyy = 0. Just as initial-value problems are associated with hyperbolic PDE. y) is harmonic The Laplace equation is arguably the most important diﬀerential equation in . the problem is independent of time. u(x. Hence. boundary value problems are associated with PDE of elliptic type. In contrast to initial-value problems. Solutions of this equation are called harmonic functions.1) where the symbol ∆ is referred to as the Laplacian. Mathematically. boundary-value problems are considerably more diﬃcult to solve.1 Show that.

1) in the rectangle Ω = {(x. It arises in an astonishing variety of mathematical and physical systems. 0) = g1 (x). There are two main modiﬁcations of the Laplace equation: the Poisson equation (a non-homogeneous Laplace equation): ∆u = f (x. electromagnetism. Thus. b) = 0   uxx + uyy = 0 u(x. The separation of variables method is most successful when the boundary conditions are homogeneous. u(x. Solving Laplace’s Equation (19. b) = g2 (x). u(a. (III)  u(0. heat conduction. ranging through ﬂuid mechanics. b) = 0 uxx + uyy = 0 u(x. u(a. and on and on. y) = u(x. y) = f2 (y). Consider (19. number theory.1) Note ﬁrst that both independent variables are spatial variables and each variable occurs in a 2nd order derivative and so we will need two boundary conditions for each variable a total of four boundary conditions. 0) = 0   . geometry.158SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS all of applied mathematics. potential theory. y) = u(x. y) and the eigenvalue problem (the Helmholtz equation): ∆u = λu. y) = f1 (y). 0) = u(x. probability. solid mechanics. where in each problem three of the four conditions are homogeneous. (I) (II)  u(a. The four problems to be solved are   uxx + uyy = 0 uxx + uyy = 0   u(0. y) = u(x. (IV )  u(0. y) = f2 (y). y) = u(a. b) = g2 (x) where 0 ≤ x ≤ a and 0 ≤ y ≤ b. u(x. y) = u(a. y) = f1 (y). 0) = u(x. 0) = g1 (x). solving the Laplace’s equation in Ω requires solving four initial boundary conditions problems. y) : 0 ≤ x ≤ a. y) = u(x. λ ∈ R. 0 ≤ y ≤ b} with the Dirichlet boundary conditions u(0. b) = 0  u(0.

0) = 0 = X(x)Y (0) =⇒ Y (0) = 0 u(x. y) = u1 (x.19 ELLIPTIC TYPE: LAPLACE’S EQUATIONS IN RECTANGULAR DOMAINS159 If we let ui (x. i = 1. y) + u4 (x. y) = u4 (x. dividing for X(x)Y (y) and subtracting from both sides. 2. The process is nearly identical in many ways to what we did when we were solving the heat equation. X(x) Y (y) The left hand side is a function of x while the right hand side is a function of y. Substituting in (19. 3. y) = 0 = X(0)Y (y) =⇒ X(0) = 0 u(a. Assuming X(x)Y (y) is nonzero. 4. That is. This results in the following two ODEs X − λX = 0 and Y + λY = 0. Y (y) X (x) =− =λ X(x) Y (y) where λ is a constant. denote the solution of each of the above problems. y) = 0 = X(a)Y (y) =⇒ X(a) = 0 u(x. As far as the boundary conditions. y) + u3 (x. we obtain X (x)Y (y) + X(x)Y (y) = 0. we will apply separation of variables to (19. we have for all 0 ≤ x ≤ a and 0 ≤ y ≤ b u(0.1). So let’s assume that the solution can be written in the form u(x. In each of the above problems.1) and ﬁnd a product solution that will satisfy the diﬀerential equation and the three homogeneous boundary conditions. y) = X(x)Y (y). We will illustrate how to ﬁnd u(x. Using the Principle of Superposition we will ﬁnd a solution to the problem and then apply the ﬁnal boundary condition to determine the value of the constant(s) that are left in the problem. Y (y) Y (y) . y) + u2 (x. This says that they must equal to a constant. y). then the solution to our original system will be u(x. we ﬁnd: X (x) Y (y) =− . y). b) = g2 (x) = X(x)Y (b). y).

the functions nπ nπ un (x. Hence. We must have B = 0 otherwise √ X(x) = 0 and √ leads to this the trivial solution. In this case. Xn (x) = sin a Now. y) = u(a.1) and the boundary conditions u(0. Thus. λ = 0 imply A = B = 0 and hence the solution is the trivial solution. The condition X(0) = 0 implies A = 0. n ∈ N. solving the equation Y + λY = 0 we obtain √ Yn (y) = an e −λn y + bn e − √ −λn y = An cosh −λn y + Bn sinh −λn y. Solving for λ we ﬁnd λn = − naπ . If λ = 0 then X(x) = Ax+B. n ∈ N. b) = g2 (x) in (19. So assume that λ = 0. The condition X(a) = 0 implies √ B sin −λa = 0. y) = n=1 Bn sin nπ nπ x sinh y . b) = g2 (x). a . a a (19. y) = Bn sin x sinh y . Now. we obtain inﬁnitely 2 many solutions given by nπ x.2) to obtain ∞ g2 (x) = n=1 Bn sinh nπ b a sin nπ x . in order to have a nontrivial solution we must have λ < 0. Using the boundary condition Y (0) = 0 we obtain An = 0 for all n ∈ N. we obtain sin −λa = 0 or −λa = nπ 2 2 where n ∈ N. If λ > 0 then B √ X(x) = Ae λx + Be− λx . 0) = 0.2) To determine the unknown constants Bn we use the boundary condition u(x.160SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS Note that X and Y are not the zero functions for otherwise u ≡ 0 and this contradicts our assumption that u is the non-trivial solution. the conditions X(0) = X(a) = 0. Now. √ √ X(x) = A cos −λx + B sin −λx. we invoke the superposition principle of linear PDE to write ∞ u(x. Since B = 0. n∈N a a satisfy (19. y) = u(x. Hence. Consider the ﬁrst equation: since X − λX = 0 the solution depends on the sign of λ. Now. in order for these solutions to satisfy the boundary value condition u(x. the conditions X(0) = X(a) = 0 imply A = √ = 0 and so u ≡ 0.

1). Assume that the solution can be written in the form u(x.19 ELLIPTIC TYPE: LAPLACE’S EQUATIONS IN RECTANGULAR DOMAINS161 Since the right-hand side is the Fourier sine series of f on the interval [0. the solution to the Laplace’s equation is given by (19. the coeﬃcients Bn are given by Bn = 2 a a g2 (x) sin 0 nπ nπ −1 x dx [sinh b ] . dividing for X(x)Y (y) and subtracting from both sides.1) with the Bn s calculated from (19. X(x) Y (y) The left hand side is a function of x while the right hand side is a function of y. we have for all 0 ≤ x ≤ a and 0 ≤ y ≤ b u(0. we ﬁnd: X (x) Y (y) =− .  u(a. 0) = u(x.3) Thus. This says that they must equal to a constant. y) = X(x)Y (y). That is. Example 19. y) = u(x. This results in the following two ODEs X − λX = 0 and Y + λY = 0. Substituting in (19. y) = f1 (y) = X(0)Y (y) u(a. y) = 0 = X(a)Y (y) =⇒ X(a) = 0 Y (y) Y (y) .3). b) = 0 Solution.2 Solve   uxx + uyy = 0 u(0. a a (19. we obtain X (x)Y (y) + X(x)Y (y) = 0. As far as the boundary conditions. X (x) Y (y) =− =λ X(x) Y (y) where λ is a constant. a]. y) = f1 (y). Assuming X(x)Y (y) is nonzero.

the condition Y (0) = Y (b) = 0 imply A = B = 0 and hence the solution is the trivial solution. Thus. Now. using the boundary condition X(a) = 0 we obtain An = 0 for all n ∈ N. y) = Bn sin nπ nπ y sinh (x − a) . So assume that λ = 0. Hence. b b However. n ∈ N. Xn (x) = An cosh nπ nπ (x − a) + Bn sinh (x − a) . this is not really suited for dealing with the boundary condition X(a) = 0. The condition Y (0) = 0 implies A = 0. √ √ Y (y) = A cos λy + B sin λy. b b Now. we obtain sin λb = 0 or λb = nπ where 2 π2 n ∈ N. n ∈ N. in order to have a nontrivial solution we must have λ > 0. n ∈ N. If λ = 0 then Y (y) = Ay +B. λ > 0 we obtain √ Xn (x) = an e λn x + bn e − √ λn x = An cosh nπ nπ x + Bn sinh x . solving the equation X − λX = 0. Since B = 0. Consider the second equation: since Y +λY = 0 the solution depends on the sign of λ. the conditions Y (0) = Y (b) = 0 imply A = √ = 0 and so u ≡ 0. we obtain inﬁnitely many solutions given by nπ y . In this case. So. We must have B = 0 otherwise Y (y) = 0 and this leads to √ √ the trivial solution. Solving for λ we ﬁnd λn = nb2 . Yn (y) = sin b Now.162SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS u(x. 0) = 0 = X(x)Y (0) =⇒ Y (0) = 0 u(x. If λ < 0 then B √ Y (y) = Ae −λy + Be− −λy . Now. n ∈ N b b . Hence. The condition Y (b) = 0 implies √ B sin λb = 0. the functions un (x. b) = 0 = X(x)Y (b) =⇒ Y (b) = 0 Note that X and Y are not the zero functions for otherwise u ≡ 0 and this contradicts our assumption that u is the non-trivial solution. let’s also notice that the following is also a solution.

19 ELLIPTIC TYPE: LAPLACE’S EQUATIONS IN RECTANGULAR DOMAINS163 satisfy (19. 0 < y < H u(0. n ∈ N. y) = u(L. 0 < x < L.4) to obtain ∞ f1 (y) = n=1 Bn sinh − nπ a b sin nπ y . Now. Using separation of variables we ﬁnd Y X =− = λ. we invoke the superposition principle of linear PDE to write ∞ u(x.5) Thus. in order for these solutions to satisfy the boundary value condition u(0. b Since the right-hand side is the Fourier sine series of f1 on the interval [0. y) = u(x. b) = 0. u(x. the coeﬃcients Bn are given by Bn = 2 b b f1 (y) sin 0 nπ y dy b sinh − nπ a b −1 .3 Solve uxx + uyy = 0.4) with the Bn s calculated from (19.1) and the boundary conditions u(a. 0 < x < L. y) = f1 (y) in (19. 0 < g < H u(x. y) = f1 (y). Solution. X Y We ﬁrst solve X − λX = 0 0<x<L X(0) = X(L) = 0 2 2 We ﬁnd λn = − nLπ and 2 Xn (x) = sin nπ x. y) = 0. the solution to the Laplace’s equation is given by (19. 0) = u(x. b b (19. b]. 0).4) To determine the unknown constants Bn we use the boundary condition u(0. L . 0) = uy (x.5) Example 19. (19. y) = n=1 Bn sin nπ nπ y sinh (x − a) . H) = f (x).

the general solution is given by ∞ u(x. L Hence. y) = n=1 Bn sin nπ nπ nπ nπ x cosh y + sinh y L L L L . L L The boundary condition Y (0) − Y (0) = 0 implies nπ An − Bn = 0. with the Bn satisfying Bn nπ nπ nπ cosh H + sinh H L L L 2 = L L f (x) sin 0 nπ x dx L . H) = f (x) we ﬁnd ∞ f (x) = n=1 Bn sin nπ nπ nπ nπ x cosh H + sinh H L L L L . Recall the Fourier sine series of f on [0. n ∈ N. nπ nπ nπ cosh y + Bn sinh y y. n ∈ N. y) = n=1 Bn sin nπ nπ nπ nπ x cosh y + sinh y L L L L . L 0 L Thus. Yn = Bn L L L Using the superposition principle and the results above we have Yn (y) = An cosh ∞ u(x. L] given by ∞ f (x) = n=1 An sin nπ x L where nπ 2 L An = f (x) sin x dx. Substituting in the condition u(x.164SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS Next we need to solve Y + λY = 0 0<y<H Y (0) − Y (0) = 0 The solution of the ODE is nπ nπ y + Bn sinh y y.

y) = u(x.e. y) = u(L. u(x.  u(0. b) = 0 Exercise 19.5 Solve uxx + uyy = 0.  u(0. Hint: Deﬁne U (x.2 Solve   uxx + uyy = 0 u(x. y) + iv(x. y) = n=0 an z n . y) = 0. Exercise 19. 0) = u(0. y) = 2xy are harmonic functions. 0 ≤ x ≤ L. y) = u(a. y) = f2 (y). y) = 2y. b) = 0 Exercise 19.1 Solve   uxx + uyy = 0 u(a. y) = 0. We say that f is holomorphic or analytic if and only if f can be expressed as a power series in z.19 ELLIPTIC TYPE: LAPLACE’S EQUATIONS IN RECTANGULAR DOMAINS165 Practice Problems Exercise 19. y) = u(x. 0) = u(x. u(x. 2 2 Exercise 19. 1) = 3 sin πx + 2x where 0 ≤ x ≤ 1 and 0 ≤ y ≤ 1. u(x. y) − 2xy.6 √ Consider a complex valued function f (z) = u(x.3 Solve   uxx + uyy = 0 u(x. y) = x2 − y 2 and u(x. i. 0) = g1 (x). y) = u(x. ∞ u(x. 0 ≤ x ≤ L. Exercise 19.4 Show that u(x. − − H H <y< 2 2 H H ≤y≤ 2 2 H H ) = f1 (x).  u(1. (a) By diﬀerentiating with respect to x and y show that . ) = f2 (x). y) where i = −1. − subject to u(0. y) + iv(x.

0) = 0. 0)| < ∞ uy (x. 0 ≤ x ≤ L. y) = 4 cos 2H Exercise 19. 0 ≤ y ≤ H subject to u(0. y) = f (y). 0 ≤ x ≤ 2. u(x. x 2 4π y .8 Solve uxx + uyy = 0. 3 . u(0. 0 ≤ y ≤ H subject to uy (x.166SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS ux = vy and uy = −vx These are known as the Cauchy-Riemann equations. 0) = 0. 0) = uy (x. H) = 0 πy . y) = u(L. x > 0.9 Solve uxx + uyy = 0. H) = 0. 3) = u(0. r r Exercise 19. u(x. 0 ≤ y ≤ 3 subject to u(x.10 Solve uxx + uyy = 0. Exercise 19. |u(x. u(2. (b) Show that ∆u = 0 and ∆v = 0. y) = sin Exercise 19. y) = 7.7 Show that Laplace’s equation in polar coordinates is given by 1 1 urr + ur + 2 uθθ = 0.

19 ELLIPTIC TYPE: LAPLACE’S EQUATIONS IN RECTANGULAR DOMAINS167

Sample Exam Questions
Exercise 19.11 Consider Laplace’s equation inside a rectangle uxx + uyy = 0, 0 ≤ x ≤ L, 0 ≤ y ≤ H subject to the boundary conditions u(0, y) = 0, u(L, y) = 0, u(x, 0)−uy (x, 0) = 0, u(x, H) = 20 sin Find the solution u(x, y). Exercise 19.12 Solve Laplace’e equation uxx + uyy = 0 in the rectangle 0 < x, y < 1 subject to the conditions u(0, y) = u(1, y) = 0, 0 < y < 1 u(x, 0) = sin (2πx), ux (x, 0) = −2π sin (2πx), 0 < x < 1. Exercise 19.13 Find the solution to Laplace’s equation on the rectangle 0 < x < 1, 0 < y < 1 with boundary conditions u(x, 0) = 0, u(x, 1) = 1 ux (0, y) = ux (1, y) = 0. Exercise 19.14 Solve Laplace’s equation on the rectangle 0 < x < a, 0 < y < b with the boundary conditions ux (0, y) = −a, ux (a, y) = 0 uy (x, 0) = b, uy (x, b) = 0. Exercise 19.15 Solve Laplace’s equation on the rectangle 0 < x < π, 0 < y < 2 with the boundary conditions u(0, y) = u(π, y) = 0 uy (x, 0) = 0, uy (x, 2) = 2 sin 3x − 5 sin 10x. πx −5 sin L 3πx . L

168SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS

20 Laplace’s Equations in Circular Regions
In the previous section we solved the Dirichlet problem for Laplace’s equation on a rectangular region. However, if the domain of the solution is a disc, an annulus, or a circular wedge, it is useful to study the two-dimensional Laplace’s equation in polar coordinates. It is well known in calculus that the cartesian coordinates (x, y) and the polar coordinates (r, θ) of a point are related by the formulas x = r cos θ and y = r sin θ
y where r = (x2 + y 2 ) 2 and tan θ = x . Using the chain rule we obtain
1

ux =ur rx + uθ θx = cos θur −

sin θ uθ r

uxx =uxr rx + uxθ θx sin θ sin θ urθ cos θ = cos θurr + 2 uθ − r r sin θ cos θ sin θ uθ − uθθ + − sin θur + cos θurθ − − r r r cos θ uy =ur ry + uθ θy = sin θur + uθ r uyy =uyr ry + uyθ θy cos θ cos θ urθ sin θ = sin θurr − 2 uθ + r r cos θ sin θ cos θ uθ + uθθ + cos θur + sin θurθ − r r r Substituting these equations into ∆u = 0 we obtain 1 1 urr + ur + 2 uθθ = 0. r r Example 20.1 Find the solution to ∆u = 0, x2 + y 2 < a2 subject to u(x, y) = f (θ), x2 + y 2 = a2 . (20.1)

20 LAPLACE’S EQUATIONS IN CIRCULAR REGIONS

169

Solution. Two important facts that play an important role in the solution of the problem must be taken into consideration, namely: (1) u(r, θ) must be bounded at r = 0. (2) u(r, θ + 2π) = u(r, θ), that is, u is periodic of period 2π. Now, we will apply the method of separation of variables to (21.1). Suppose that a solution u(r, θ) of (21.1) can be written in the form u(r, θ) = R(r)Θ(θ). Substituting in (21.1) we obtain 1 1 R (r)Θ(θ) + R (r)Θ(θ) + 2 R(r)Θ (θ) = 0 r r Dividing by RΘ (under the assumption that RΘ = 0) we obtian R (r) Θ (θ) R (r) = −r2 −r . Θ(θ) R(r) R(r) The left-hand side is independent of r whereas the right-hand side is independent of θ so that there is a constant λ such that − R (r) R (r) Θ (θ) = r2 +r = λ. Θ(θ) R(r) R(r)

This results in the following ODEs Θ (θ) + λΘ(θ) = 0 and r2 R (r) + rR (r) − λR(r) = 0 (20.3) The second equation is known as Euler’s equation. Both of these equations are easily solvable. We only have to add the appropriate boundary conditions. By (2) above we must have Θ(θ + 2π) = Θ(θ), ∀θ ∈ R. In particular, we have Θ(0) = Θ(2π) and Θ (0) = Θ (2π). The periodicity of Θ implies that λ = n2 and Θ must be of the form Θn (θ) = An cos nθ + Bn sin nθ, n = 0, 1, 2 · · · (20.2)

170SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS The equation in R is of Euler type and its solution must be of the form R(r) = rα . Since λ = n2 , the corresponding characteristic equation is α(α − 1)rα + αrα − n2 rα = 0. Solving this equation we ﬁnd α = ±n. Hence, we let Rn (r) = Cn rn + Dn r−n , n ∈ N. For n = 0, R = 1 is a solution. To ﬁnd a second solution, we solve the equation r2 R + rR = 0. This can be done by dividing through by r and using the substitution S = R to obtain rS + S = 0. Solving this by noting that the left-hand side is just c c (rS) we ﬁnd S = r . Hence, R = r and this implies R(r) = C ln r. Thus, R = 1 and R = ln r form a couple of linearly independent solutions of (20.3) and so a general solution is given by R0 (r) = C0 + D0 ln r. By assumption (1), u(r, θ) must be bounded at r = 0, and so does Rn . Since r−n and ln r are unbounded at r = 0, we must set D0 = Dn = 0. In this case, the solutions to Euler’s equation are given by Rn (r) = Cn rn , n = 0, 1, 2, · · · . Using the superposition principle, and combining the results obtained above, we ﬁnd ∞ u(r, θ) = C0 +
n=1

rn (An cos nθ + Bn sin nθ).

Now, using the boundary condition u(a, θ) = f (θ) we can write

f (θ) = C0 +
n=1

(an An cos nθ + an Bn sin nθ)

which is usually written in a more convenient equivalent form by a0 f (θ) = + (an cos nθ + bn sin nθ). 2 n=1

20 LAPLACE’S EQUATIONS IN CIRCULAR REGIONS

171

It is obvious that an and bn are the Fourier coeﬃcients, and therefore can be determined by the formulas an = and 1 π

f (θ) cos nθdθ, n = 0, 1, · · ·
0

1 2π f (θ) sin nθdθ, n = 1, 2, · · · . π 0 Finally, the general solution to our problem is given by bn =

u(r, θ) = C0 +
n=1

rn (An cos nθ + Bn sin nθ)

where
2π a0 1 = f (θ)dθ 2 2π 0 2π an 1 An = n = n f (θ) cos nθdθ, n = 1, 2, · · · a a π 0 2π bn 1 Bn = n = n f (θ) sin nθdθ, n = 1, 2, · · · a a π 0

C0 =

Example 20.2 Solve ∆u = 0, 0 ≤ θ < 2π, 1 ≤ r ≤ 2 subject to u(1, θ) = u(2, θ) = sin θ, 0 ≤ θ < 2π. Solution. Use separation of variables. First, solving for Θ(θ)), we see that in order to ensure that the solution is 2π−periodic in θ, the eigenvalues are λ = n2 . When solving the equation for R(r), we do NOT need to throw out solutions which are not bounded as r → 0. This is because we are working in the annulus where r is bounded away from 0 and ∞. Therefore, we obtain the general solution

u(r, θ) = (C0 + C1 ln r) +
n=1

[(Cn rn + Dn r−n ) cos nθ + (An rn + Bn r−n ) sin nθ].

B1 = 2 . θ) = cos 9θ − cos 3θ. θ) = 0. 3 3 9 You may assume that the solution remains bounded as r → 0. Hence. comparing coeﬃcients we must have C0 Cn + Dn An + Bn A1 + B1 n Cn 2 + Dn 2−n An 2n + Bn 2−n 2A1 + 2−1 B1 =0 =0 =0 n = 1 =1 =0 =0 n = 1 =0 1 Solving these equations we ﬁnd C0 = Cn = Dn = 0. θ) = 1 3 r+ 2 r sin θ Example 20. and 3 An = Bn = 0 for n = 1. the solution to the problem is u(r. Solution. Separating the variables we obtain the eigenvalue problem Θ (θ) + λΘ(θ) = 0 Θ (0) = Θ π 3 = 0. uθ (r. A1 = 3 .3 Solve Laplace’s equation inside a 60◦ wedge of radius a subject to the boundary conditions π 1 1 uθ (r. . ) = 0.172SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS But C0 + n=1 n=1 ∞ ∞ [(Cn + Dn ) cos nθ + (An + Bn ) sin nθ] = sin θ and ∞ ∞ C0 + n=1 n=1 [(Cn 2n + Dn 2−n ) cos nθ + (An 2n + Bn 2−n ) sin nθ] = sin θ Hence. u(a.

20 LAPLACE’S EQUATIONS IN CIRCULAR REGIONS 173 As above. 2. θ) = n=0 Cn r3n cos 3nθ. 1. 1. Thus. n = 0. · · · The corresponding solutions of the radial problem are Rn (r) = An r3n + Bn r3 n . · · · . 2. 1. Hence. 1 r 9 a 3 cos 9θ − cos 3θ . the angular solution is Θn (θ) = cos 3nθ. ∞ π 3 = 0 implies un (r. · · · . θ) = 1 r 3 a 9 1 1 cos 9θ − cos 3θ 3 9 1 3 and 0 otherwise. · · · Using the boundary condition u(a. The condition Θ λn = (3n)2 . n = 0. The condition Θ (0) = 0 implies A = 0. n = 0. n = 0. Thus. 1. because of periodicity we expect the solution to be of the form √ √ Θ(θ) = A cos λθ + B sin λθ. To obtain a solution that remains bounded as r → 0 we take Bn = 0. θ) = we obtain C1 a3 = − 1 and C3 a9 = 9 u(a. 2.

θ) = 3 sin 5θ. 2.5 (a) Using Euler’s formula from complex analysis eit = cos t + i sin t show that 1 cos t = (eit + e−it ).4 Consider u(r. θ) = C0 + n=1 ∞ rn (An cos nθ + Bn sin nθ) with 2π 1 a0 = f (φ)dφ 2 2π 0 2π an 1 An = n = n f (φ) cos nφdφ. 2 . Exercise 20. 2. Exercise 20.2 Solve the Laplace’s equation in the upper half of the unit disk with u(1. n = 1. Exercise 20.3 Solve the Laplace’s equation in the unit disk with ur (1. θ) = 2π 2π ∞ f (φ) 1 + 2 0 n=1 r a n cos n(θ − φ) dφ. θ) = π − θ. Exercise 20.1 Solve the Laplace’s equation in the unit disk with u(1.174SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS Practice Problems Exercise 20. · · · a a π 0 2π bn 1 Bn = n = n f (φ) sin nφdφ. n = 1. θ) = 2 cos 2θ. · · · a a π 0 C0 = Using the trigonometric identity cos a cos b + sin a sin b = cos (a − b) show that 1 u(r.

− 2ar cos (θ − φ) + r2 This is known as the Poisson formula in polar coordinates. − π ≤ θ ≤ π. Using these concepts. r r (c) Let q1 = a ei(θ−φ) and q2 = a e−i(θ−φ) .20 LAPLACE’S EQUATIONS IN CIRCULAR REGIONS √ where i = −1. (b) Show that ∞ 175 1+2 n=1 r a n ∞ cos n(θ − φ) = 1 + n=1 r a n ∞ e in(θ−φ) + n=1 r a n e−in(θ−φ) . show that |q1 | < 1 and |q2 | < 1. a2 − 2ar cos (θ − φ) + r2 Exercise 20.7 Show that a2 − r 2 u(r. x2 + y 2 < 1 subject to u(1. (b) Show that ∞ 1+2 n=1 r a n cos n(θ − φ) = a2 − r 2 . θ) = θ. θ) = 2π 2π 0 a2 f (φ) dφ.6 (a)Show that ∞ n=1 r a r a n n ein(θ−φ) = rei(θ−φ) a − rei(θ−φ) re−i(θ−φ) a − re−i(θ−φ) and ∞ e−in(θ−φ) = n=1 Hint: Each sum is a geoemtric series with a ratio less than 1 in absolute value so that these series converges. Exercise 20. Exercise 20. .8 Solve uxx + uyy = 0. It is deﬁned in complex analysis that 1 the absolute value of a complex number z = x+iy is given by |z| = (x2 +y 2 ) 2 .

t) depends on r and t only can be describe by the one-dimensional wave equation in polar coordinates 1 utt = c2 (urr + ur ). 0 < r < a.1 below. Exercise 20. t > 0 r with initial condition u(a. 0) = f (r). ut (r.10 Cartesian coordinates and cylindrical coordinates are shown in Figure 22. t > 0 and boundary conditions u(r. t) = 0. 0 < r < a. Figure 22. c2 T R rR (b) Show that λ < 0. 0) = g(r).9 The vibrations of a symmetric circular membrane where the displacement u(r.176SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS Exercise 20. t) = R(r)T (t) leads to the equation 1 1R 1T = R + = λ.1 . (a) Show that the assumption u(r.

z = z. y = r sin θ. (b) Show that 1 1 uxx + uyy + uzz = urr + ur + 2 uθθ + uzz . r r 177 .20 LAPLACE’S EQUATIONS IN CIRCULAR REGIONS (a) Show that x = r cos θ.

Exercise 20. y) ∈ Ω ∪ ∂Ω. Let u be harmonic in Ω = {(x.y)∈Ω u(x. y). y). y) ≤ max u(x.y)∈∂Ω where ∂Ω denotes the boundary of Ω. y) (ii) min(x. y) ≤ u(x. y) and u2 |∂Ω = g3 (x. y) deﬁned in a domain Ω satisﬁes the inequality min u(x. y) > 0 for all (x.13 Let u1 (x. y) for all (x. θ) = sin2 θ. y) be harmonic functions on a smooth domain Ω such that u1 |∂Ω = g1 (x.15 Solve the Dirichlet problem ∆u = 0.12 Let u be harmonic in Ω = {(x.178SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS Sample Exam Questions Exercise 20.11 An important result about harmonic functions is the so-called the maximum principle which states: Any harmonic function u(x.y)∈∂Ω (x. y) = 2 − x for all (x.y)∈Ω u(x. y) ∈ ∂Ω. 0 ≤ r < a. y) where g1 and g2 are continuous functions satisfying (x. . y) = 1 + 3x for all (x. y) ∈ ∂Ω. y) : x2 + y 2 < 1} and satisﬁes u(x.14 Show that rn cos (nθ) and rn sin (nθ) satisfy Laplace’s equation in polar coordinates. Exercise 20. Exercise 20.y)∈∂Ω max g1 (x. Show that u(x. − π ≤ θ ≤ π u(a. y) ∈ Ω ∪ ∂Ω (x. y) ∈ Ω. y) < min g1 (x.y)∈∂Ω Prove that u1 (x. y) without solving ∆u = 0. y) : x2 + y 2 < 1} and satisﬁes u(x. ∀(x. y) < u2 (x. (x. y) and u2 (x. Determine (i) max(x. Exercise 20.

θ) = ln 2 + 4 cos 3θ.16 Solve Laplace’s equation uxx + uyy = 0 outside a circular disk (r ≥ a) subject to the boundary condition u(a.20 LAPLACE’S EQUATIONS IN CIRCULAR REGIONS Exercise 20. You may assume that the solution remains bounded as r → ∞. 179 .

180SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS .

The fundamental ones are the Laplace and the Fourier transforms. where the equations are transformed to algebraic equations. Examples of evolutions equations are the heat equation and the wave equation. Examples of stationary equations are the Laplace’s equations and Poisson equations. where the transformation leads to the decrease of the number of independent variables. PDEs in two variables are thus reduced to ODEs. In contrast. and is used especially to solve linear ODEs with constant coeﬃcients. • This simple equation is solved by purely algebraic manipulations.The Laplace Transform Solutions for PDEs If in a partial diﬀerential equation the time t is one of the independent variables of the searched-for function. 21 Essentials of the Laplace Transform Laplace transform has been introduced in an ODE course. when the equation involves only spatial independent variables then the equation is called a stationary equation. • The solution of the simple equation is transformed back to obtain the so181 . This idea can be easily extended to PDEs. There are classes of methods that can be used for solving the initial value or initial boundary problems for evolution equations. we say that the PDE is an evolution equation. In this section we review the Laplace transform and its properties. In this chapter we will just consider the Laplace transform. The process of solution consists of three main steps: • The given “hard” problem is transformed into a “simple” equation. Laplace transform is yet another operational tool for solving constant coeﬃcients linear diﬀerential equations. We refer to these methods as the methods of integral transforms.

So what types of functions possess Laplace transforms. When the improper integral is convergent then we say that the function f (t) possesses a Laplace transform. The third step is made easier by tables. 0 But 0 T e−(s−a)t dt = 1−e−(s−a)T s−a if s = a if s = a.1 In this section we introduce the concept of Laplace transform and discuss some of its properties. . what type of functions guarantees a convergent improper integral.182 THE LAPLACE TRANSFORM SOLUTIONS FOR PDES lution of the given problem. whose role is similar to that of integral tables in integration. Then the Laplace transform of f. if it exists.1 Figure 21. that is.1 Find the Laplace transform. is deﬁned by the following equation T ∞ L[f (t)] = F (s) = lim T →∞ f (t)e−st dt = 0 0 f (t)e−st dt The integral which deﬁnes a Laplace transform is an improper integral. The above procedure can be summarized by Figure 21. An improper integral may converge or diverge. which is denoted by L[f (t)] or by F (s). In this way the Laplace transformation reduces the problem of solving a differential equation to an algebraic problem. Example 21. Let f (t) be deﬁned for t ≥ 0. of each of the following functions (a) f (t) = eat (b) f (t) = 1 (c) f (t) = t (d) f (t) = et 2 Solution. (a) Using the deﬁnition of Laplace transform we see that ∞ T L[eat ] = 0 e−(s−a)t dt = lim T T →∞ e−(s−a)t dt. depending on the integrand. The Laplace transform is deﬁned in the following way.

L[eat ] = F (s) = 1 .1 below) the integral on the left is also ∞ ∞ divergent. s > 0. Now. A function f that satisﬁes |f (t)| ≤ M eat . Since the integral on the right is divergent. if s > 0 then 0 et(t−s) dt ≥ s dt. s > a. we ﬁnd ∞ T L[1] = 0 e −st dt = lim T →∞ 0 1 e−st dt = . To answer this question we introduce few mathematical concepts. ∞ 2 If s ≤ 0 then t2 −st ≥ 0 so that et −st ≥ 1 and this implies that 0 et −st dt ≥ ∞ dt. s > 0. s2 (d) Again using the deﬁnition of Laplace transform we ﬁnd L[e ] = 0 2 t2 ∞ et 2 −st dt.21 ESSENTIALS OF THE LAPLACE TRANSFORM For the improper integral to converge we need s > a. By the same reasoning 2 the integral on the left is divergent. by the comparison theorem 0 of improper integrals (see Theorem 23.e. t≥C (21. Below is a sketch of a piecewise continuous . In this case. A function f is called piecewise continuous on an interval if the interval can be broken into a ﬁnite number of subintervals on which the function is continuous on each open subinterval (i. s−a 183 (b) In a similar way to what was done in part (a). the subinterval without its endpoints) and has a ﬁnite limit at the endpoints (jump discontinuities and no vertical asymptotes) of each subinterval.1) is said to be a function with an exponential order at inﬁnity. This shows that the function f (t) = et does not possess a Laplace transform The above example raises the question of what class or classes of functions possess a Laplace transform. s (c) We have ∞ L[t] = 0 te−st dt = − te−st e−st − 2 s s ∞ = 0 1 .

n n (a) Since et = ∞ t ≥ t .184 function. . Note that the two conditions above are suﬃcient. Theorem 21.2 Show that the following functions are piecewise continuous and of exponential order at inﬁnity for t ≥ 0 (a) f (t) = tn (b) f (t) = tn sin at Solution.1 Suppose that f (t) is piecewise continuous on t ≥ 0 and has an exponential order at inﬁnity with |f (t)| ≤ M eat for t ≥ C. tn ≤ n!et . tn is piecewise continuous and n=0 n! n! of exponential order at inﬁnity. The next theorem shows that any linear combination of functions in PE is also in PE. Then the Laplace transform ∞ F (s) = 0 f (t)e−st dt exists as long as s > a. The same is true for the product of two functions in PE. for F (s) to exist. Example 21. tn sin at is piecewise continuous and of exponential order at inﬁnity The following is an existence result of Laplace transform. A function deﬁned for t ≥ 0 is said to be piecewise continuous on the inﬁnite interval if it is piecewise continuous on 0 ≤ t ≤ T for all T > 0. Hence. we will denote the class of all piecewise continuous functions with exponential order at inﬁnity by PE. THE LAPLACE TRANSFORM SOLUTIONS FOR PDES Note that a piecewise continuous function is a function that has a ﬁnite number of breaks in it and doesn’t blow up to inﬁnity anywhere. but not necessary. (b) Since |tn sin at| ≤ n!et . In what follows.

s > a. Then f (t) = g(t) for all t ≥ 0 where both functions are continuous. how do we invert the transform. t ≥ C1 and |g(t)| ≤ M2 ea1 t . The following result on uniqueness provides a possible answer. we have that L[eat ] = s−a . This result establishes a one-to-one correspondence between the set PE and its Laplace transforms. Hence. L−1 s−1 = et . The above theorem states that if f (t) is continuous and has a Laplace transform F (s). We next discuss the problem of how to determine the function f (t) if F (s) is given.3 Let f (t) and g(t) be two elements in PE with Laplace transforms F (s) and G(s) such that F (s) = G(s) for some s > a. we can inspect tables of Laplace transforms of . t ≥ 0 . Solution. In particular. then there is no other function that has the same Laplace transform. To ﬁnd L−1 [F (s)]. we can now oﬃcially deﬁne the inverse Laplace transform as follows: For a piecewise continuous function f of exponential order at inﬁnity whose Laplace transform is F. t ≥ C2 . Symbolically f (t) = L−1 [F (s)] ⇐⇒ F (s) = L[f (t)]. Theorem 21. That is. (ii) The function h(t) = f (t)g(t) is an element of PE.21 ESSENTIALS OF THE LAPLACE TRANSFORM Theorem 21. s > 1. Moreover L[αf (t) + βg(t)] = αL[f (t)] + βL[g(t)].3 1 Find L−1 s−1 . 1 From Example 23. Alternatively. we call f the inverse Laplace transform of F and write f = L−1 [F (s)].1(a). s > 1.2 Suppose that f (t) and g(t) are two elements of PE with |f (t)| ≤ M1 ea1 t . Example 21. for 1 1 a = 1 we ﬁnd that L[et ] = s−1 . 185 (i) For any constants α and β the function αf (t) + βg(t) is also a member of PE. the following theorem asserts that the Laplace transform of a member in PE is unique. With the above theorem.

4 Consider the two functions f (t) = H(t)H(3 − t) and g(t) = H(t) − H(t − 3). Solution. the uniqueness of the inverse Laplace transform is not assured. s > 0. (a) We have f (t) = and g(t) = 1.4 Given two Laplace transforms F (s) and G(s) then L−1 [aF (s) + bG(s)] = aL−1 [F (s)] + bL−1 [G(s)] for any constants a and b.186 THE LAPLACE TRANSFORM SOLUTIONS FOR PDES known functions to ﬁnd a particular f (t) that yields the given F (s). t>3 1. 0 ≤ t ≤ 3 0. where H is the Heaviside function deﬁned by H(t) = (a) Are the two functions identical? (b) Show that L[f (t)] = L[g(t). t ≥ 0 0. Theorem 21. However. s Thus. The following example addresses the uniqueness issue. t≥3 1. t < 0 Since f (3) = 1 and g(3) = 0 then f and g are not identical. (b) We have 3 L[f (t)] = L[g(t)] = 0 e−st dt = 1 − e−3s . When the function f (t) is not continuous. they are equal on the interval(s) where they are both continuous The inverse Laplace transform possesses a linear property as indicated in the following result. both functions f (t) and g(t) have the same Laplace transform even though they are not identical. Example 21. 0 ≤ t < 3 0. .

Then (i) f ∗ g = g ∗ f (Commutative Law) (ii) (f ∗ g) ∗ k = f ∗ (g ∗ k) (Associative Law) (iii) f ∗ (c1 g + c2 k) = c1 f ∗ g + c2 f ∗ k (Distributive Law) Example 21.6 Express the solution to the initial value problem y + αy = g(t). Using integration by parts twice we arrive at (f ∗ g)(t) = = = 1 2 sin sds e (sin s − cos s) e−t 1 + 2 (sin t − cos t) 2 t −(t−s) e 0 −(t−s) t 0 Next.5 Find f ∗ g where f (t) = e−t and g(t) = sin t. Theorem 21. we state several properties of convolution product. .21 ESSENTIALS OF THE LAPLACE TRANSFORM 187 Convolution integrals are useful when ﬁnding the inverse Laplace transform of products.5 Let f (t). They are deﬁned as follows: The convolution of two scalar piecewise continuous functions f (t) and g(t) deﬁned for t ≥ 0 is the integral t (f ∗ g)(t) = 0 f (t − s)g(s)ds. known as the Convolution Theorem. Example 21. which resemble those of ordinary product. g(t). Solution. and k(t) be three piecewise continuous scalar functions deﬁned for t ≥ 0 and c1 and c2 are arbitrary constants. y(0) = y0 in terms of a convolution integral. Solving this initial value problem by the method of integrating factor we ﬁnd t y(t) = e−αt y0 + 0 e−α(t−s) g(s)ds = e−αt y0 + (e−αt ∗ g)(t) The following theorem. provides a way for ﬁnding the Laplace transform of a convolution integral and also ﬁnding the inverse Laplace transform of a product. Solution.

8 Solve the initial value problem 4y + y = g(t). and of exponential order at inﬁnity then L[(f ∗ g)(t)] = L[f (t)]L[g(t)] = F (s)G(s). Thus. y (0) = −7 Solution. (s2 + a2 )2 . 1 a So. Take the Laplace transform of all the terms and plug in the initial conditions to obtain 4(s2 Y (s) − 3s + 7) + Y (s) = G(s) or (4s2 + 1)Y (s) − 12s + 28 = G(s).188 THE LAPLACE TRANSFORM SOLUTIONS FOR PDES Theorem 21. (f ∗ g)(t) = L−1 [F (s)G(s)]. F (s) = G(s) = Thus. in this case we have. Example 21. Note that P (s) = 1 s 2 + a2 1 s 2 + a2 1 s2 +a2 1 . Solving for Y (s) we ﬁnd .7 Use the convolution theorem to ﬁnd the inverse Laplace transform of P (s) = Solution. sin (at − as) sin (as)ds = 0 1 (sin (at) − at cos (at)) 2a3 Example 21.6 If f (t) and g(t) are piecewise continuous for t ≥ 0. (f ∗ g)(t) = 1 a2 t so that f (t) = g(t) = sin (at). y(0) = 3.

2 So. . H is the Heaviside function and δ is the Dirac delta function as deﬁned in Section 0. once we decide on a g(t) all we need to do is to evaluate the integral and we’ll have the solution We conclude this section with the following table of Laplace transform pairs. 12s − 28 G(s) + 1 2+ 1 4 s 4 s2 + 4 4 3s s2 + ( 1 2 2 − 1 2 2 7 2+ 1 2 s 2 1 1 2 + G(s) 2+ 4 s 2 1 2 2 y(t) = 3 cos t 2 − 7 sin t 2 1 + 2 t sin 0 s g(t − s)ds.21 ESSENTIALS OF THE LAPLACE TRANSFORM 189 Y (s) = = Hence.

t < 0 1 s. s>α (s−α)2 +ω 2 s−α . s>α (s−α)n+1 ω . (s2 +ω 2 )2 sin ωt − ωt cos ωt] s>0 s>0 1 [sin ωt 2ω 3 1 . with f (t) continuous and |f (t)| ≤ M eat f (t). s>0 s2 +ω 2 s . s > a e e−αs s . · · · eαt sin (ωt) eαt cos (ωt) f (t − α)H(t − α). 0} + 1 Table L . 2. s > |ω| s2 −ω 2 F (s − α). t 0 f (u)du. s>0 s2 +ω 2 ω . 0} + 1 s2 F (s) − sf (0) − f (0) s > max{a. s>0 sn+1 s . s > α n! . α ≥ 0 tf (t) t 2ω n! . s>0 tn . 0} + 1 sn F (s) − sn−1 f (0) − · · · -sf (n−2) (0) − f (n−1) (0) s > max{a. with f (t) continuous and |f (t)| ≤ M eat f (n) (t). with f (n−1) (t) continuous and |f (n) (t)| ≤ M eat sF (s) − f (0) s > max{a. s>α s−α ω . s > |ω| s2 −ω 2 s . s>α (s−α)2 +ω 2 −αs F (s). 2. α ≥ 0 with |f (t)| ≤ M eat H(t − α). with |f (t)| ≤ M eat s > max{a. 0} + 1 F (s) s . s > α + a 1 s−α . n = 1. (s2 +ω 2 )2 f (t). with |f (t)| ≤ M eat eαt H(t) eαt tn . · · · eαt sin (ωt) cos (ωt) sinh (ωt) cosh (ωt) eαt f (t). t ≥ 0 0. n = 1.190 f(t) H(t) = THE LAPLACE TRANSFORM SOLUTIONS FOR PDES F(s) 1. s>0 -F (s) s .

Exercise 21. Exercise 21. Exercise 21. If the integral con- ∞ t dt 0 1+t2 converges. ∞ 1 dt 0 1+t2 converges. ﬁnd L[t − 5].7 Using the deﬁnition. if it exists. ﬁnd L[f (t)]. if it exists. f (t) = 0. 0≤t<1 t − 1. give its value.21 ESSENTIALS OF THE LAPLACE TRANSFORM 191 Practice Problems Exercise 21. if it exists. Exercise 21. ﬁnd L[e(t−1) ]. If the Laplace transform exists then ﬁnd the domain of F (s). Exercise 21. ﬁnd L[(t − 2)2 ].1 Determine whether the integral verges. if it exists. if it exists. give its value. If the Laplace transform exists then ﬁnd the domain of F (s). t≥1 . ﬁnd L[e3t ]. If the integral con- ∞ −t e 0 cos (e−t )dt converges. If the Laplace transform exists then ﬁnd the domain of F (s). Exercise 21.5 Using the deﬁnition. If the integral Exercise 21. If the Laplace transform exists then ﬁnd the domain of F (s).8 Using the deﬁnition.2 Determine whether the integral verges.3 Determine whether the integral converges.4 Using the deﬁnition. give its value. If the Laplace transform exists then ﬁnd the domain of F (s).6 2 Using the deﬁnition.

16 Use Table L to ﬁnd L[2et + 5]. .11 For s > 0 and n a positive integer evaluate the limits (a) limt→0 tn e−st (b) limt→∞ tn e−st Exercise 21.10 Let n be a positive integer. t e dt = − s s Exercise 21. 2 s−2 . 1 ≤ t < 2 f (t) =  0.192 THE LAPLACE TRANSFORM SOLUTIONS FOR PDES Exercise 21.17 Use Table L to ﬁnd L[e3t−3 H(t − 1)]. ﬁnd L[f (t)].9 Using the deﬁnition.  0≤t<1  0. Using integration by parts establish the reduction formula tn e−st n n −st + tn−1 e−st dt. If the Laplace transform exists then ﬁnd the domain of F (s).13 3 Find L−1 s−2 . Find the domain of F (s).15 2 Find L−1 s+2 + 1 s+1 . Exercise 21. t ≥ 2. Exercise 21.12 Use the linearity property of Laplace transform to ﬁnd L[5e−7t + t + 2e2t ]. s > 0. Exercise 21.14 Find L−1 − s22 + Exercise 21. t − 1. if it exists. Exercise 21. Exercise 21.18 Use Table L to ﬁnd L[sin2 ωt]. Exercise 21.

21 ESSENTIALS OF THE LAPLACE TRANSFORM Exercise 21.21 Use Table L to ﬁnd L[e4t (t2 + 3t + 5)]. Exercise 21. s−3 193 . 4 ].23 5 Use Table L to ﬁnd L−1 [ (s−3)4 ].20 Use Table L to ﬁnd L[e2t cos 3t]. Exercise 21.19 Use Table L to ﬁnd L[sin 3t cos 3t].22 Use Table L to ﬁnd L−1 [ s210 + +25 Exercise 21. Exercise 21.

Exercise 21. 1 ≤ t < 3 g(t) =  0. 0 ≤ t < 1 12. Exercise 21.26 Using the partial fraction decomposition ﬁnd L−1 12 (s−3)(s+1) .29 Consider the functions f (t) = et and g(t) = e−2t . y (0) = 0. (a) By directly evaluating the integral. (b) By computing L−1 [F (s)G(s)] where F (s) = L[f (t)] and G(s) = L[g(t)].30 Consider the functions f (t) = sin t and g(t) = cos t. . s−9 Exercise 21. (b) By computing L−1 [F (s)G(s)] where F (s) = L[f (t)] and G(s) = L[g(t)]. y(0) = 2 where   0. t ≥ 0. Exercise 21.27 Use Laplace transform technique to solve the initial value problem y + 4y = g(t). t≥3 Exercise 21. y(0) = 0. Compute f ∗ g in two diﬀerent ways. Exercise 21.194 THE LAPLACE TRANSFORM SOLUTIONS FOR PDES Sample Exam Questions Exercise 21.25 Using the partial fraction decomposition ﬁnd L−1 Exercise 21.24 −2s Use Table L to ﬁnd L−1 [ e ]. 24e−5s s2 −9 . Compute f ∗ g in two diﬀerent ways. (a) By directly evaluating the integral.31 Compute t ∗ t ∗ t. t ≥ 0.28 Use Laplace transform technique to solve the initial value problem y − 4y = e3t .

21 ESSENTIALS OF THE LAPLACE TRANSFORM Exercise 21.33 Compute t ∗ e−t ∗ et . Exercise 21.32 Compute H(t) ∗ e−t ∗ e−2t . 195 .

t) = 1. Let us assume that at time t = 0 the concentration is 0 and on the boundary x = 0. x > 0. constant unit concentration of the contaminant is kept for t > 0.  u(0. 0) = 0. 0). The transformation will be done with respect to the time variable t ≥ 0. s) = s2 U (x. 0 Likewise. for example L(ut )(x. t > 0   u(x. s). for example. t)| < ∞ Find u(x. The behaviour of this problem is described by the following mathematical model   ut − uxx = 0 . In particular we deﬁne the Laplace transform of u(x.196 THE LAPLACE TRANSFORM SOLUTIONS FOR PDES 22 Solving PDEs Using Laplace Transform The same idea for solving linear ODEs using Laplace transform can be exploited when solving PDEs for functions in two variables u = u(x. s) − su(x. the spatial variable x will be treated as a parameter unaﬀected by this transform. s) − u(x. t) = Uxx (x. t) be the concentration of a chemical contaminant dissolved in a liquid on a half-inﬁnte domain x > 0.1 Let u(x. that is. t). t) by the formula ∞ L(u(x. t) = 0 ux (x.   |u(x. 0) and L(utt )(x. applying the Laplace transform to a PDE in two variables x and t we obtain an ODE in the variable x and with the parameter s. Thus. we have Luxx (x. τ )e −sτ ∂ dτ = ∂x ∞ u(x. τ )e−sτ dτ = Ux (x. ∞ Lux (x. 0) − ut (x. t). s). t)) = U (x. s) = 0 u(x. The spatial derivatives remain unchanged. Example 22. . The time derivatives are transformed in the same way as in the case of functions in one variable. τ )e−sτ dτ. t) = sU (x.

t) = f (t). Applying Laplace transform to both sides of the equation we obtain sU (x. 197 This is a second order linear ODE in the variable x and positive parameter s. 0) − Uxx (x. t > 0   u(x. s) = A(s)e− √ sx √ + B(s)e sx .  u(0.22 SOLVING PDES USING LAPLACE TRANSFORM Solution. s Thus. Next. t) = L−1 1 −√sx e . t)| < ∞ . x > 0. Since U (x. s) = 0. we must have B(s) = 0 and in this case we obtain √ U (x. s √ 1 − sx e s One can use a software package to ﬁnd the expression for L−1 Example 22. s) − sU (x. s) = e− sx . s) = A(s)e− sx . s) − u(x. we apply Laplace transform to the boundary condition obtaining 1 U (0. s) = L(1) = .   |u(x. Its general solution is U (x.2 Solve the following initial boundary value problem   ut − uxx = 0 . s) = 0 or Uxx (x. s) is bounded in both variables. u(x. 0) = 0. s This leads to A(s) = 1 s and the transformed solution becomes 1 √ U (x.

0) = ut (x. t) = f (t). t > 0    u(x. t) = L−1 F (s)e− It can be shown that L−1 (e− Hence. using Theorem 21.198 THE LAPLACE TRANSFORM SOLUTIONS FOR PDES Solution. x > 0. Applying Laplace transform to both sides of the equation we obtain s2 U (x. t)| < ∞ Solution. 0) − ut (x. s) = 0. s) = 0 or c2 Uxx (x. This is a second order linear ODE in the variable x and positive parameter s. F (s) = Lf (t). s) = A(s)e− c x + B(s)e c x . Its general solution is U (x. s) = A(s)e− c x . t) = 0 x 4π(t − s)3 e− 4(t−s) f (s)ds x2 Example 22.    |u(x. s) − s2 U (x. s) − su(x. Following the argument of the previous example we ﬁnd U (x. s s s . s) is bounded. Since U (x. 4πt3 √ sx = f ∗ L−1 (e− √ sx ). u(x. Thus. s) = F (s)e− √ sx . we must have B(s) = 0 and in this case we obtain U (x. 0) = 0.3 Solve the wave equation  utt − c2 uxx = 0 . 0) − c2 Uxx (x. u(0.6 we can write u(x. t √ sx x2 x )= √ e− 4t .

t) = L−1 F (s)e− c s = H t − x s 199 x x f t− c c Remark 22.22 SOLVING PDES USING LAPLACE TRANSFORM Next. Thus. s) = F (s)e− c x .1 Laplace transforms are useful in solving parabolic and some hyperbolic PDEs. . we apply Laplace transform to the boundary condition obtaining U (0. This leads to A(s) = F (s) and the transformed solution becomes U (x. They are not in general useful in solving elliptic PDEs. u(x. s) = L(f (t)) = F (s).

 |u(x. 0) = sin x. 0) = e−5x . 0) = 2 sin πx + 3 sin 2πx. Exercise 22. 0) = 0. t) = 0 Hint: Method of integrating factor of ODEs. t > 0 u(x. t > 0 u(x. x > 0. x > 0.  u(0.  u(0.5 Solve by Laplace transform   ut + ux = t . Hint: A particular solution of a second order ODE must be found using the method of variation of parameters. t > 0  ut + ux = 0 u(x. t > 0 u(x. 0) = sin x.200 THE LAPLACE TRANSFORM SOLUTIONS FOR PDES Practice Problems Exercise 22. t) = u(1. x > 0. t) = t2 .1 Solve by Laplace transform  .3 Solve ut = 4uxx u(0. t) = 0 u(x. x > 0. Exercise 22.2 Solve by Laplace transform   ut + ux = −u .4 Solve by Laplace transform   ut − ux = u . t) = 0 Exercise 22. t)| < ∞ Exercise 22.  u(0.

6 Solve by Laplace transform   xut + ux = 0 . ut (x. t > 0 u(x. t > 0    u(x. t)| < ∞. t) = 0. x > 0. 0 ≤ x ≤ π.    |u(x. x > 0. x > 0. 0) = 2 sin x. ux (0. t) = f (t).7 Solve by Laplace transform  utt − c2 uxx = 0 .    |u(x.  u(0. t) = t Exercise 22. 0) = 0.  u(0. 0) = 0. Exercise 22. y > 0  u(x.10 Solve by Laplace transform  utt = c2 uxx .9 Solve by Laplace transform  uxy = 1 . 0) = ut (x. t)| < ∞ Exercise 22. x > 0. 0) = 0. t > 0 u(0.8 Solve by Laplace transform utt − 9uxx = 0. 0) = ut (x. t) = sin x. y) = y + 1. u(0.22 SOLVING PDES USING LAPLACE TRANSFORM Exercise 22. 201 . Exercise 22. u(x. t > 0    u(x. 0) = 1. t) = u(π. 0) = 0.

0) = sin x.  u(0. t) = 0. x u(x. 0) = 3 sin πx. 0 ≤ x ≤ 2. u(x. t) = u(2. 0) = T.13 Solve by Laplace transform ut − 3uxx = 0.14 Solve by Laplace transform ut − 4uxx = 0. t) = 0. t > 0   u(x. t) = u(2. t) = 0. t > 0  ut + ux = u u(x. 0) = 0.12 Solve by Laplace transform   ut − c2 uxx = 0 .   |u(x.15 Solve by Laplace transform utt − 4uxx = 0.  u(0. t > 0 ux (0. 0 ≤ x ≤ π. t) = u(π.202 THE LAPLACE TRANSFORM SOLUTIONS FOR PDES Sample Exam Questions Exercise 22. 0 ≤ x ≤ 2. x > 0. x > 0. t)| < ∞ Exercise 22.11 Solve by Laplace transform  . t) = 0 Exercise 22. 0) = 5 sin (πx) Exercise 22. t > 0 u(0. . t) = 0. u(x. ut (x. 0) = 40 cos 2 Exercise 22. t > 0 u(0.

the Laplace transfom. Fourier series are designed to solve boundary value problems on bounded intervals. It is often convenient to recast this series in complex form by means of Euler formula eix = cos x + i sin x. In the context of Fourier analysis. signal processing. including both ordinary and partial diﬀerential equations. control theory. quantum mechanics. The extension of Fourier methods to the entire real line leads naturally to the Fourier transform. 23 Complex Version of Fourier Series We have seen in Section 15 that a 2L−periodic function f : R → R that is piecewise smooth on [−L. In this chapter. to name but a few. The Fourier transform is of fundamental importance in a broad range of applications. L] can be expanded in a Fourier series nπ nπ a0 + an cos x + bn sin x f (x) = 2 L L n=1 at all points of continuity of f. the so-called Fourier transform. It follows from this formula that 203 ∞ . and probability. we consider a second fundamental class of integral transform methods. this is referred to as the real form of the Fourier series. an extremely powerful mathematical tool for the analysis of non-periodic functions.The Fourier Transform Solutions for PDEs In the previous chapter we discussed one class of integral transform methods.

204 THE FOURIER TRANSFORM SOLUTIONS FOR PDES eix + e−ix = 2 cos x and eix − e−ix = 2i sin x or cos x = eix +e−ix 2 and sin x = eix −e−ix . In order to ﬁnd these formulas.2) That is. 2i Hence the Fourier expansion of f can be rewritten as a0 f (x) = + an 2 n=1 +bn e inπx L ∞ e inπx L + e− 2 inπx L − e− 2i ∞ inπx L f (x) = −∞ cn e inπx L (23. (23. an and bn can be easily found once we have formulas for cn .1) where c0 = a0 2 and for n ∈ N we have cn = c−n = an − ibn 2 an + ibn . we need to evaluate the following integral L e −L inπx L e− imπx L L dx = −L e i(n−m)πx L dx i(n−m)πx L L e L i(n − m)π −L iL =− [cos [(n − m)π] + i sin [(n − m)π] (n − m)π − cos [−(n − m)π] − i sin [−(n − m)π]] =0 = . 2 It follows that if n ∈ N then an = cn + c−n and bn = i(cn − c−n ).

if we multiply (23. 2L −L Example 23.23 COMPLEX VERSION OF FOURIER SERIES if n = m. Using integration by parts and the fact that eiπ = e−iπ = −1 we ﬁnd π 1 xe−inx dx cn = 2π −π π π 1 ix −inx i = e − e−inx dx 2π n n −π −π 1 iπ −inπ iπ inπ = e + e 2π n n 1 1 1 −inπ e − 2 einπ + 2 2π n n 1 π 1 (−1)n i = 2i (−1)n + (0) = 2π n 2π n Remark 23. For example. n ∈ N −π ≤x≤π . extended to be periodic of period 2π. Now.1 It is often the case that the complex form of the Fourier series is far simpler to calculate than the real form. · · · . ±1.1) by e− the last result we ﬁnd L −L inπx L and integrate from −L to L and apply inπx L f (x)e− dx = 2Lcn which yields the formula for coeﬃcients of the complex form of the Fourier series: L inπx 1 cn = f (x)e− L dx. n = 0. One can then use (23. Solution.1 Find the complex Fourier coeﬃcients of the function f (x) = x. ±2. If n = m then L 205 e −L inπx L e− inπx L dx = 2L. the Fourier coeﬃcients of the real form of the previous function are given by 2 an = (cn + c−n ) = 0 and bn = i(cn − c−n ) = n (−1)n+1 .2) to ﬁnd the real form of the Fourier series.

recover the real Fourier series representation of f. Exercise 23.5 Find the complex Fourier series of the 2π−periodic function deﬁned f (x) = 1 0<x<T 0 T < x < 2π Exercise 23.6 Let f (x) = x2 . (b) Find the complex Fourier series representation of f.3 Find the complex Fourier series of the 2π−periodic function f (x) = eax over the interval (−π.1 Find the complex Fourier coeﬃcients of the function f (x) = x.2 Let −1≤x≤1   0 −π < x < −π 2 1 −π < x < π f (x) = 2 2  0 π<x<π be 2π−periodic. Exercise 23. Exercise 23. Find its complex series representation. (a) Calculate the complex Fourier series representation of f. π). be of period 1. . Exercise 23. be 2π−periodic.206 THE FOURIER TRANSFORM SOLUTIONS FOR PDES Practice Problems Exercise 23. bn and cn . − π < x < π. − 1 < x < 2 . π).7 1 Let f (x) = sin nπx. (b) Using the complex Fourier series found in (a).4 Find the complex Fourier series of the 2π−periodic function f (x) = sin x over the interval (−π. extended to be periodic of period 2. 2 (a) Calculate the coeﬃcients an . Exercise 23.

(b) Find the complex Fourier series representation of f. bn and cn .23 COMPLEX VERSION OF FOURIER SERIES Exercise 23. n = 0. Using these deﬁnitions show that an = 2Re(cn ) and bn = −2Im(cn ). n = 1. · · · .8 Let f (x) = 2 − x. be of period 2. − 2 < x < 2. . 1. · · · and bn . Find an . Exercise 23. 2.10 Recall that any complex number z can be written as z = Re(z) + iIm(z) where Re(z) is called the real part of z and Im(z) is called the imaginary part. 207 Exercise 23. 2. (a) Calculate the coeﬃcients an . The complex conjugate of z is the complex number z = Re(z) − iIm(z).9 Suppose that the coeﬃcients cn of the complex Fourier series are given by cn = 2 iπn 0 if |n| is odd if |n| is even.

11 Suppose that cn = Find an and bn . (a) Write f (x) explicitly.14 1 If z is a complex number we deﬁne sin z = 2 (eiz − e−iz ). (c) Determine bn for n ∈ N. Exercise 23.12 Find the complex Fourier series of the function f (x) = ex on [−2. Exercise 23. 2].15 Find cn for the 2π−periodic function f (x) = 1 if s ≤ x ≤ s + h 0 elsewhere in [−π.208 THE FOURIER TRANSFORM SOLUTIONS FOR PDES Sample Exam Questions Exercise 23. (d) Determine c0 and cn for n ∈ N. (b) Determine a0 and an for n ∈ N. Find the complex form of the Fourier series for sin 3x without evaluating any integrals. Exercise 23. Exercise 23.13 Consider the wave form i [e−inT 2πn T 2π − 1] if n = 0 if n = 0. π] . What is the period of f.

n Now. of period 2πL. Let ξ ∈ R. There are many times when one would like to divide a function which is not periodic into a superposition of sines and cosines.2) ˆ The function f is called the Fourier transform of f. Like the Laplace transform. πL]. To start with. . Assuming integration and summation can be interchanged we ﬁnd πL ∞ πL f (x)e −πL −iξx dx = n=−∞ cn −πL e−iξx e inx L dx. the Fourier transform is often an eﬀective tool in ﬁnding explicit solutions to diﬀerential equations.24 TWO DIMENSIONAL FOURIER TRANSFORMS 209 24 Two Dimensional Fourier Transforms One of the problems with the theory of Fourier series discussed so far is that it applies only to periodic functions. Multiply both sides of (24. This function can be extended to a periodic function. still denoted by f . say to f (ξ). We will use the notation ˆ F[f (x)] = f (ξ). we ﬁnd ∞ ˆ f (ξ) = −∞ f (x)e−iξx dx (24. letting ξ = L we ﬁnd ˆ n = f L πL f (x)e− −πL inx L dx = 2πLcn . Hence. The Fourier transform is the tool often used for this purpose.1) where 1 cn = 2πL πL f (x)e− −πL inx L . From the previous section we can ﬁnd the complex Fourier series of f to be ∞ f (x) = n=−∞ cn e inx L (24.1) by e−iξx and then integrate both sides from −πL to πL. ˆ It can be shown that the RHS converges. as L → ∞. let f : R → R be a piecewise continuous function that vanishes outside an interval of the form [−πL.

3) is called the Fourier inversion formula. Equations (24.3) Equation (24. We have ∞ ∞ e−ax if x ≥ 0 0 if x < 0 ˆ f (ξ) = −∞ ∞ f (t)e−iξx dx = e 0 −ax−iξx e−ax e−iξx dx 0 −x(a+iξ) ∞ = e dx = −(a + iξ) 0 1 = a + iξ The following theorem lists the basic properties of Fourier transform .210 THE FOURIER TRANSFORM SOLUTIONS FOR PDES Hence. Solution. If we make use of Euler’s formula. Example 24. (24. L f (x) = f 2πL n=−∞ L In the limit as L → ∞. and our formula becomes 1 ˆ F −1 [f (ξ)] = f (x) = 2π ∞ ∞ ˆ f (ξ)eiξx dξ −∞ (24.1 Find the Fourier transform of the function f (x) deﬁned by f (x) = for some a > 0. f (x) = 1 2π i ˆ f (ξ) cos ξxdξ + 2π −∞ ∞ ∞ ˆ f (ξ) sin ξxdξ −∞ a superposition of sines and cosines of various frequencies. we can write the Fourier inversion formula in terms of sines and cosines.3) allow one to pass back and forth between a given function and its representation as a superposition of oscillations of various frequencies.2) and (24.1) can be written in the form 1 ˆ n e inx . it can be shown that this last sum approaches an improper integral.

We have u(ξ) = ˆ −∞ ∞ e−αx e−iξx dx. be piecewise continuous functions.2 2 Determine the Fourier transform of the Gaussian u(x) = e−αx . ∞ ∞ 1 ˆ (7) Parseval’s Relation: −∞ |f (x)|2 dx = 2π −∞ |f (ξ)|2 dξ. Solution.24 TWO DIMENSIONAL FOURIER TRANSFORMS 211 Theorem 24. ∞ ˆ (4) Continuity: If −∞ |f (x)|dx < ∞ then f is continuous in ξ. ˆ (9) Multiplication by xn : F[xn f (x)] = in f (n) (ξ). x (3) Scaling: F[f α ] = αF[f (αx)]. α > 0. Then we have the following properties: (1) Linearity: F[αf (x) + βg(x)] = αF[f (x)] + βF[g(x)]. π (10) Gaussians: F[e−αx ] = α e− 4α . (5) Diﬀerentiation: F[f (n) (x)] = (iξ)n F[f (x)). g.1 Let f. where α and β are arbitrary numbers. 2 If we diﬀerentiate this relation with respect to the variable ξ and then integrate by parts we obtain ∞ u (ξ) = − i ˆ i = 2α iξ = 2α −∞ ∞ −∞ ∞ −∞ xe−αx e−iξx dx d −αx2 −iξx (e )e dx dx 2 2 (e−αx )e−iξx dx = − ξ u(ξ) ˆ 2α ξ Thus we have arrived at the ODE u (ξ) = − 2α y (ξ) whose general solution ˆ ˆ has the form ξ2 u(ξ) = Ce− 4α ˆ . x 1 (6) Integration: F 0 f (s)ds = − iξ F[f (x)]. (12) Convolution: F[(f ∗ g)(x)] = F[f (x)] · F[g(x)]. 1 (11) Product: F[(f (x)g(x)] = 2π F[f (x)] ∗ F[g(x)]. 2 ξ2 Example 24. (8) Duality: F[F[f (x)]] = 2πf (−x). (2) Shifting: F[f (x − α)] = e−αξ F[f (x)].

ˆ ˆ (7’) Convolution: F −1 [f ∗ g (ξ)] = 2π(f g)(x). 1 (5’) Gaussians: F −1 [e−αξ ] = √4πα e− 4α .4 Prove F[F[f (x)]] = 2πf (−x). 2πf (−x) = −∞ ∞ 1 2π ∞ ˆ f (ξ)eiξx dξ −∞ ˆ ˆ f (ξ)e−iξx dξ = F[f (ξ)] = F[F[f (x)]] The following theorem lists the properties of inverse Fourier transform Theorem 24. ˆ (3’) Multiplication by ξ n : F −1 [ξ n f (ξ)] = (−i)n f (n) (x). Solution. We have f (x) = Thus.3 Prove ˆ F[f (−x)] = f (−ξ). Using a change of variables we ﬁnd ∞ ∞ F[f (−x)] = −∞ f (−x)e−iξx dx = −∞ ˆ f (x)eiξx dx = f (−ξ) Example 24.212 Since THE FOURIER TRANSFORM SOLUTIONS FOR PDES ∞ u(0) = ˆ −∞ e−αx dx = π − ξ2 e 4α α π =C α we ﬁnd u(ξ) = ˆ Example 24.2 Let f and g be piecewise continuous functions. ˆ ˆ (1’) Linearity: F −1 [αf (ξ) + βˆ(ξ)] = αF −1 [f (ξ)] + βF −1 [ˆ(ξ)]. −iξα −1 −iξα ˆ (4’) Multiplication by e : F [e f (ξ)] = f (x − α). Solution. 2 x2 . g g −1 ˆ(n) n (2’) Derivatives: F [f (ξ)] = (−ix) f (x). ˆ g (6’) Product: F −1 [f (ξ)ˆ(ξ)] = f (x) ∗ g(x).

2) with the constant √1 or 2π in front of the integral.1 It is important to mention that there exists no established convention how to deﬁne the Fourier transform. we can meet an equivalent 1 deﬁnition of (24.24 TWO DIMENSIONAL FOURIER TRANSFORMS 213 Remark 24. . The reader should keep this fact in mind while working with various sources or using the transformation tables. In literature. 2π There also exist deﬁnitions with positive sign in the exponent.

0) = g(x). 0) = f (x). Exercise 24. L) = 1 if −a < x < a 0 otherwise Exercise 24. .3 Obtain the transformed problem when applying the Fourier transform with respect to the spatial variable to the equation and both initial conditions utt = c2 uxx .214 THE FOURIER TRANSFORM SOLUTIONS FOR PDES Practice Problems Exercise 24. 0 < y < L u(x. t > 0 u(x. 0) = 0 u(x. 0) = f (x) ut (x.4 Obtain the transformed problem when applying the Fourier transform with respect to the spatial variable to the equation and both initial conditions ∆u = uxx + uyy = 0.2 Obtain the transformed problem when applying the Fourier transform with respect to the spatial variable to the equation and initial condition ut + cux = 0 u(x. x ∈ R. Exercise 24. Exercise 24. where α > 0.5 Find the Fourier transform of f (x) = e−|x|α . x ∈ R.1 Find the Fourier transform of the function f (x) = 1 if −1 ≤ x ≤ 1 0 otherwise.

24 TWO DIMENSIONAL FOURIER TRANSFORMS Exercise 24.8 Prove ˆ F[f (x − α)] = e−iξα f (ξ).9 Prove ˆ F[eiαx f (x)] = f (x − α). 1 + ix . 1 = 2πeξ H(−ξ). Exercise 24.10 Prove the following 1 ˆ ˆ F[cos (αx)f (x)] = [f (ξ + α) + f (ξ − α)] 2 1 ˆ ˆ F[sin (αx)f (x)] = [f (ξ + α) − f (ξ − α)] 2 215 1 1 + iξ 1 if x ≥ 0 0 otherwise.7 Prove that F Exercise 24. Exercise 24.6 Prove that F[e−x H(x)] = where H(x) = Exercise 24.

12 Find the Fourier transform of f (x) = 1 − |x| for −1 ≤ x ≤ 1 and 0 otherwise.14 ξ2 ˆ Find the inverse Fourier transform of f (ξ) = e− 2 . the Fourier transform of   −1 −a < x < 0 1 0<x<a f (x) =  0 otherwise Exercise 24.13 Find. Exercise 24.216 THE FOURIER TRANSFORM SOLUTIONS FOR PDES Sample Exam Questions Exercise 24. Exercise 24. . using the deﬁnition.15 1 Find F −1 a+iξ . Exercise 24.11 Prove ˆ F[f (x)] = (iξ)f (ξ).

We solve this ODE to obtain the transformed solution u which can be converted to the original solution u ˆ by means of the inverse Fourier transform. the Fourier transform is applied to the spatial variable on the whole real line. First Order Transport Equation Consider the initial value problem ut + cux = 0 u(x. Performing the Fourier transˆ form on both the PDE and the initial condition. t) = f (ξ)e−iξct . Let u(ξ. 0) = f (x). t) be the Fourier transform of u in x. x ∈ R. we apply Fourier transforms in solving various PDE problems. In this section. The Fourier transform will be applied to the spatial variable x while the variable t remains ﬁxed.25 APPLICATIONS OF FOURIER TRANSFORMS TO PDES 217 25 Applications of Fourier Transforms to PDEs Fourier transform is a useful tool for solving diﬀerential equations. The PDE in the two variables x and t passes under the Fourier transform to an ODE in the t−variable. ˆ Solution of the ODE gives ˆ u(ξ. Second Order Wave Equation Consider the two dimensional wave equation utt = c2 uxx . t) = F −1 [u(ξ. Contrary to Laplace transform. we reduce the PDE into an ODE in t ∂u ˆ + iξcˆ = 0 u ∂t ˆ u(ξ. u(x. 0) = f (ξ). ˆ Thus. t)] = f (x − ct) which is exactly the same as we obtained by using the method of characteristics. t > 0 . We illustrate these ideas in the examples below. which usually uses the time variable.

0) = g (ξ). we ﬁnd the inverse transforms of Φ and Ψ φ(x) = 1 1 f (x) + 2 c 1 1 f (x) − 2 c x g(s)ds 0 x ψ(x) = g(s)ds . 0) = f (x) ut (x. ˆ ˆ General solution to the ODE is u(ξ. But Φ(ξ) = 1 1 ˆ f (ξ) − g (ξ) ˆ 2 iξc 1 1 ˆ f (ξ) + g (ξ) . Again. 0) = f (ξ) ˆ ut (ξ. by performing the Fourier transform of u in x. t) = φ(x − ct) + ψ(x + ct) ˆ ˆ where φ = Φ and ψ = Ψ. Performing the inverse transformation and making use of the translation theorem. we reduce the PDE problem into an ODE problem in the variable t: ∂ 2u ˆ = −c2 ξ 2 u ˆ 2 ∂t ˆ u(ξ. x−ct . 0) = g(x).218 THE FOURIER TRANSFORM SOLUTIONS FOR PDES u(x. we get the general solution u(x. t) = Φ(ξ)e−iξct + Ψ(ξ)eiξct ˆ where Φ and Ψ are two arbitrary functions of ξ. t) = [f (x − ct) + f (x + ct)] + 2 2c x+ct g(s)ds. 0 Application of the translation property then yields directly the D’Alambert solution 1 1 u(x. ˆ 2 iξc Ψ(ξ) = By using the integration property.

L) = 1 if −a < x < a 0 otherwise 2 Performing Fourier Transform in x for the PDE we obtain the second order PDE in y uyy = ξ 2 u. 0) = 0 u(x. we consider the heat equation ut = kuxx . ˆ . x ∈ R. ˆ ˆ The general solution is given by u(ξ. x ∈ R. y) = A(ξ) sinh ξy + B(ξ) cosh ξy. t) = f (ξ)e−kξ t . t) =f (x) ∗ F −1 ]e−kξ t ] x2 1 =f (x) ∗ √ e− 4kt 4πkt ∞ (x−s)2 1 =√ f (s)e− 4kt ds 4πkt −∞ Laplace’s Equation in 2D Consider the problem ∆u = uxx + uyy = 0. ˆ Application of the convolution theorem yields u(x. 0 < y < L u(x. 219 Performing Fourier Transform in x for the PDE and the initial condition. t > 0 u(x.25 APPLICATIONS OF FOURIER TRANSFORMS TO PDES Second Order Heat Equation Next. 0) = f (x). ˆ Treating ξ as a parameter. we obtain ∂u ˆ = −kξ 2 u ˆ ∂t ˆ u(ξ. we obtain the solution to the above ODE problem 2 ˆ u(ξ. 0) = f (ξ).

u(ξ. ξ sinh ξL Noting that the integrand is an even function in ξ. L) = ˆ −∞ a u(x. L)e−iξx dx a = −a e−ξx dx = −a cos ξxdx = Hence. y) = 1 2π ∞ −∞ 2 sin ξa sinh ξyeiξx dξ. y) = sinh ξy cos ξxdξ 2π −∞ ξ sinh ξL . Using the second ˆ boundary condition we ﬁnd ∞ u(ξ. 2 sin ξa ξ 2 sin ξa ξ A(ξ) sinh ξL = and this implies A(ξ) = Thus. ξ sinh ξL 2 sin ξa sinh ξy. ξ sinh ξL Taking inverse Fourier transform we ﬁnd u(x.220 THE FOURIER TRANSFORM SOLUTIONS FOR PDES Using the boundary condition u(ξ. y) = ˆ 2 sin ξa . 0) = 0 we ﬁnd B(ξ) = 0. we can simplify a little to to obtain ∞ 2 sin ξa 1 u(x.

6 Solve the Laplace’s equation in the half plane uxx + uyy = 0. x2 x2 Exercise 25. . Exercise 25. by using Fourier transform ut + cux = 0 u(x.5 Prove that ∞ e−|ξ|y eiξx dξ = −∞ x2 2y .4 Use Fourier transform to solve the heat equation ut = uxx + u. by using Fourier transform ut = kuxx − αu. 0 < y < ∞ subject to the boundary condition u(x. 0) = e− 4 . 0) = f (x).2 Solve. Exercise 25. + y2 Exercise 25. x ∈ R u(x. 0) = 1 if x ≥ 0 0 otherwise. y)| < ∞. −∞<x<∞< t>0 u(x.25 APPLICATIONS OF FOURIER TRANSFORMS TO PDES 221 Practice Problems Exercise 25. 0) = f (x). Exercise 25. 0) = e− γ . − ∞ < x < ∞. |u(x.1 Solve.3 Solve the heat equation ut = kuxx subject to the initial condition u(x.

0) = e−x using the Fourier transform. Exercise 25.9 Solve the initial value problem ut = kuxx u(x.222 THE FOURIER TRANSFORM SOLUTIONS FOR PDES Exercise 25. . 0) = e−x using the Fourier transform.7 Use Fourier transform to ﬁnd the transformed equation of utt + (α + β)ut + αβu = c2 uxx where α. β > 0.8 Solve the initial value problem ut + 3ux = 0 u(x. Exercise 25.

0) = e−x using the Fourier transform. .10 Solve the initial value problem ut = kuxx u(x. y) = 0.12 Solve. by using Fourier transform ∆u = 0 uy (x.25 APPLICATIONS OF FOURIER TRANSFORMS TO PDES 223 Sample Exam Questions Exercise 25. Exercise 25. 0) = f (x) x2 +y 2 →∞ 2 lim u(x. Exercise 25. 0) = x2 using the Fourier transform.11 Solve the initial value problem ut + cux = 0 u(x.

224 THE FOURIER TRANSFORM SOLUTIONS FOR PDES .

y. z) = z2 . z) = z2 . y. v) = 9(u (u2 +5v)2 = −9u+5v)2 . z) = x2 − 20yz 3 − 1+16y2 . y) = cos x ex y−5y (x2 − 15y 2 ) 2 +5v)−9u(2u) 2 +45v (e) fu (u. y) = x2 sin x ex y−5y +cos x ex y−5y (2xy). fy (x. y) = − 5x−3y2 (x2 + 1 225 . fy (x. t) = 2t − 4 s− 7 . y) = √y 28 (b) fx (x. y. z) = −2 x sin y z3 (g) fx (x. t) = 7t6 ln (s2 ) − 27 s 7 t4 2 3 2 3 2 3 4 4 4 4 (d) fx (x.Answers and Solutions Section 1 1. 1 2 −1 2 2x + 5 5x−3y 2 3y (x2 +ln (5x − 3y 2 ))− 2 . y.3 3 1. z) = 2xy + 43. fz (x. y. fz (x. fy (x.2 3 (a) fx (x. fy (x. fy (x. ft (s. y) = 4x3 . z) = −30y 2 z 2 3 7 (c) fs (s.4 2 2 ∂z = t2 est sin (s2 t) + 2stest cos (s2 t) ∂s 2 2 ∂z = 2stest sin (s2 t) + s2 est cos (s2 t) ∂t 1.1 (a) y 3 cos (xy) 2 (b) ex y (2y + 4x2 y 2 ) (c) 0 1. v) = (u−45u 2 2 +5v) (u2 sin y x cos y (f) fx (x. y) = ln (5x − 3y 2 )) 1. y.5 u is the depedent variable whereas x and y are the independent variables. fv (u.

By the change of variable u = −x we ﬁnd 0 0 a f (u)du = − −a a f (−u)du = 0 f (u)du. By the change of variable u = −x we ﬁnd 0 0 a f (u)du = − −a a f (−u)du = − 0 f (u)du. the result follows.9 utt = − sin uxx = − sin u vdx + uv dx. the result follows.10 utt = sin x sinh t uxx = − sin x sinh t .7 We have a 0 a f (x)dx = −a −a f (x)dx + 0 f (x)dx.6 We have ANSWERS AND SOLUTIONS a 0 a f (x)dx = −a −a f (x)dx + 0 f (x)dx. Hence. Integrate both sides to obtain uv = Now subtract 1. 1. Hence.8 By the product rule of derivatives we have (uv) = u v + uv . 1. . 1.226 1. u vdx from both sides to obtain the desired result. x x sin sin t t .

Also. n=1 2. Then |fN (x)−f (x)| = xN > 0. n=1 (b) Suppose the contrary. {fn }∞ converges pointwise to the zero function on R. n]rightarrow∞ n+1 Applying the squeeze rule for sequences. 1. In particular. Let = 1 .1 (a) For all 0 ≤ x < 1 we have limn→∞ fn (x) = limn→∞ xn = 0. Hence. Hence. we have x x2 nx + x2 = lim + lim 2 = 0 lim fn (x) = lim n→∞ n n→∞ n n→∞ n→∞ n2 Thus.12 (a) We have sup{|un (x. limn→∞ fn (1) = 1.5) N < x < 1. 1 = 0. n+1 n+1 fn (x) = 0 . t) − 1| : x ∈ R} = en t n 2 Section 2 2. the given sequence does not converge uniformly. the sequence {fn }∞ converges pointwise to f.227 1. Then there exists a positive integer N 2 such that for all n ≥ N we have |fn (x) − f (x)| < for all x ∈ [0. we obtain lim √ lim n]rightarrow∞ 1 1 ≤ fn (x) ≤ √ . 2. 0) − 1| : x ∈ R} = (b) We have sup{|un (x.5 = which is a contradiction. we have |fN (x) − f (x)| < 1 1 2 1 2 for all x ∈ [0.3 For every real number x.2 For every real number x. 1]. 1]. we have −√ Moreover. Choose (0.11 2 sup{ sinh t } 1 n 1 sup{| sin nx| : x ∈ R} = n .

4 First of all. 2 xN < 1 . 1). {fn }∞ is not pointwise convergent on [0. n→∞ lim fn (1) = 1. 1]. For x = 0 we have fn (0) = 1 for all n in N. So. {fn }∞ converges pointwise to the zero function on R. But ln x < 0 when 0 < x < 1. Then for n ≥ N xn |x|n 1 1 x− −x = < ≤ < . (b) Since limn→∞ fn (x) = 1 for all x ∈ [0. let = 1 and suppose that the convergence is uniform. n=1 2. the 2 . 2 In particular.5 For − π ≤ x < 0 and 0 < x ≤ 2 n→∞ π 2 we have lim (cos x)n = 0. π 2 2. Therefore. n n n N Thus.6 (a) Let > 0 be given. the sequence {fn }∞ converges n=1 pointwise to f (x) = 1. However. it follows that limn→∞ fn (x) = 0 for 0 < x < 1 Finally. the given sequence converges uniformly (and pointwise) to the function f (x) = x. observe that fn (0) = 0 for every n in N. 1). To see this. Therefore. the convergence is not uniform. fn (1) = n2 for all n. {fn }∞ converges n=1 pointwise to f (x) = 0 if − π ≤ x < 0 and 0 < x ≤ 2 1 if x = 0. Let N be a positive integer such that N > 1 . Then there is a 2 positive integer N such that for n ≥ N we have 1 |1 − xn−1 − 1| = |x|n−1 < .228 ANSWERS AND SOLUTIONS for all x in R. n=1 2. So the sequence {fn (0)}∞ is constant and converges to zero. Now suppose 0 < x < 1 then n=1 n2 xn = n2 en ln x . 1) and xN = 1 2 which contradicts 1 for all x ∈ [0. Thus. Hence. if we let n = N + 1 we must have xN < But x = 1 2 1 N ∈ [0.

|fn (x) − | = 2 2 4n 2(2n + sin x 3 Since limn→∞ 4n =) we can ﬁnd a positive integer N such that if n ≥ N then 3 < .229 convergence is not uniform. .7 (a) The pointwise limit is   0 if 0 ≤ x < 1 1 if x = 1 f (x) =  2 1 if 1 < x ≤ 2 (b) The convergence cannot be uniform because if it were f would have to be continuous. 7]. 2. Note that 2 cos x − sin2 x 3 1 ≤ . 2 2 2. 7 7 lim fn xdx = 2 2 n→∞ lim fn xdx = 2 1 5 dx = . Thus.9 We have proved earlier that this sequence converges pointwise to the discontinuous function f (x) = 0 if − π ≤ x < 0 and 0 < x ≤ 2 1 if x = 0 π 2 Therefore. 2 4n This shows that fn → (b) We have 7 n→∞ 1 2 uniformly on R and also on [2. 2.8 (a) Let > 0 be given. uniform convergence cannot occur for this given sequence.10 (a) Using the squeeze rule we ﬁnd n→∞ lim sup{|fn (x)| : 2 ≤ x ≤ 5} = 0. 2. for n ≥ N and all x ∈ R we have 4n 1 3 |fn (x) − | ≤ < .

230 ANSWERS AND SOLUTIONS Thus.3 y(t) = 3 sin t − 3.1 y = 2 (1 − e−t ). {fn }∞ converges uniformly to the zero function. 3. 3.5 y(t) = Ce− sin t − 3. Section 3.11 y(t) = t ln |t| + 7t.2 y(t) = 3t−1 9 + e−2t + Ce−3t . .4 y(t) = + C. n=1 (b) We have 5 n→∞ 5 lim fn (x)dx = 2 2 0dx = 0. 3 cos t t 3.10 u(x. 2 3.7 p(t) = 2 and g(t) = 2t + 3. 3.8 y0 = y(0) = −1 and g(t) = 2et + cos t + sin t. t 1 (3 sin (3t) 13 + 2 cos (3t)) + Ce−2t .6 α = −2. Suppose ﬁrst that a = λ. y) = f (bx − ay)e 3. 3. 3. 3. Then y + ay = be−at and the corresponding general solution is y(t) = bte−at + Ce−at Thus.9 1. 3. 1 3. limt→∞ y(t) = limt→∞ ( ebt + eC ) at at = limt→∞ aebat = 0 − c (ax+by) a2 +b2 .12 Since p(t) = a we ﬁnd µ(t) = eat .

3. y0 = 2 and n = 3. 12t2 3. y) = F (y)e−3x + G(x) where F (y) = . 4.13 y(t) = (−tet + et )−1 .11 α = 1 .9 y(t) = tan (t + π) = tan t.10 y(t) = 3−e−2t . 4.7 y(t) = e1−t − 1. 4.15 y(t) = tSi(t) + (3 − Si(1))t.3 y(t) = Ct2 + 4.4 y(t) = 4. Section 4 4.1 y(t) = 3 t2 e 2 +C . 3. 1+Ce4t 5 − 4 cos (2t). 4.14 y(t) = t2 4 t t −3+2+ 1 . 4.12 u(x. 4. f (y)dy. 1 3 .2 y(t) = Cet 2 −2t 4. 4. −8t+1 4.5 y(t) = 2Ce4t .6 y(t) = − (−2 cos t + 4). t→∞ b −λt e + Ce−at a−λ lim y(t) = 0.8 y(t) = √ 2 . 3+e−2t 1 2 4. suppose that a = λ then y(t) = Thus.231 Now.

232 ANSWERS AND SOLUTIONS 4.10 y(t) = −e 2 (t+π) (3 cos 2 + sin 2 ).4 y − y − 2y = 0.14 3y 2 y + cos y + 2t = 0. 1 2t t 5. . 5. 5.11 y(t) = yh (t) + yp (t) = e 2 t (c1 cos 5. 5.3 y(t) = −2e− √ 2 t 2 . 5. 5. 5.9 y(t) = 3e−t cos t + 2e−t sin t.12 y(t) = yh (t) + yp (t) = c1 e(−2− 5. limt→−∞ y(t) = 0 and limt→∞ y(t) = ∞.13 y 2 + cos y + cos t + t2 2 = 2.2 y(t) = −2 2e(−2− 2)t +2 2e(−2+ 2)t .13 y = Ax4 + Bx4 ln x. y(2) = 0. limt→−∞ y(t) = −∞ and limt→∞ y(t) = 0. Section 5 5. √ √ √ √ 5. limt→−∞ y(t) = −∞ and limt→∞ y(t) = 0.14 y = x−1 (A cos 3 ln x + B sin 3 ln x). 5. 6)t + c2 e(−2+ 6)t − t2 − 5 t − 9. √ 1 √ 3 t 2 √ + c2 sin √ 3 t) 2 + 6 73 cos 3t − 16 73 sin 3t.5 y(t) = e 3 −1 (1 − t).7 y(0) = 2 and y (0) = −2. 4. 4.1 y(t) = et − 2e3t .15 The ODE is not separable. 2 √ √ 5.6 y(t) = e− 5 (t − 1). t t 5.8 y(t) = c1 e3t + c2 te3t .

3 uss + utt = 0.15 (a) λn = n2 . yn (x) = sin nx.4 (a) Order 3.19 y(t) = 17 t e 15 1 10 1 cos (2t) + 5 sin (2t) + 5tet 3 20 t + 1 e−2t − 1 t − 1 − 6 2 4 sin 2t − 1 20 cos 2t. n = 1. · · · . homogeneous (c) Order 2. 5.16 We consider ﬁrst the cases (a) and (b). 6. L L (ky (x)) y(x)dx + 0 0 λy 2 (x)dx = 0. Section 6 6. . 2. Use integration by parts in the ﬁrst integral [ky L (x)y(x)]0 L L − 0 k(y (x)) dx + 0 2 λy 2 (x)dx = 0. nonlinear (b) Order 1. 6. 3. We solve the above equation for λ and obtain λ= L 0 k(y (x))2 dx L 0 y 2 (x)dx ≥ 0. · · · . · · · . Multiply the equation by y (x) and integrate in x from 0 to L. 2 2 1 (c) λn = π n − 2 .233 5. linear.17 y(t) = 2e 2 .1 (a) ODE (b) PDE (c) ODE. we repeat the above argument but by integrating from −L to L.2 uss = 0. The boundary term vanishes because of the boundary conditions. n = 1. n = 1. yn (x) = cos π n − 1 x. 1 π 2 π (b) λn = n − 2 L and yn = sin L n − 1 x . 2 5.18 y(t) = c1 + c2 et − 5. 2. 5. For the case (c). 6. 2.

homogeneous. z)uy +J(x. linear.5 (a) Linear. homogeneous. The inhomogeneity is − sin y. homogeneous. (d) Nonlinear. y.15 uw = u. second order. 6. (d) Nonlinear. inhomogeneous. (b) Linear. nonhomogeneous (d) Order 2. (c) PDE. y. y.10 uww = 0. quasilinear. nonlinear. homogeneous. y. z)uxz +F (x. 6. 6. z)uxx +B(x.7 (a) PDE. linear. order 3. (b) PDE. z)u = L(x. second order. (e) Order 2. 6. (b) Linear. The inhomogeneity is f (x. inhomogeneous. y. z). y. second order. linear.9 (a) Order 3. y. z)uyz + G(x. fourth order. 6. 6. The nonlinear term is uux . y. (b) Order 1. order 4. The nonlinear terms are ux uxxy and uuy . . nonlinear. z)ux +I(x. (c) Nonlinear. y. (e) Linear. linear. linear. (d) ODE. nonlinear. t). y. 2 6. (c) Nonlinear. ANSWERS AND SOLUTIONS 6. y. z)uz +K(x. inhomogeneous.6 (a) Linear.12 uvw = 0. second order.8 A(x. nonhomogeneous. (e) PDE.13 us = 0. linear.14 us = 1 . y. nonhomogeneous. homogeneous. 6. z)uyy +E(x. 6. order 2. order 2. second order.234 linear. order 3. 6. z)uzz H(x. linear.11 uvw = 0. z)uxy +C(x. (f) PDE. (c) Order 4.

x x (e y + xex ) + (x + 1)y − xe2 y x = 0 and 7. y) = y f (t)dt. 7.4 We have ux +uy +2u = e−2y cos (x − y)−2e−2y sin (x − y)−e−2y cos (x − y)+ 2e−2y sin (x − y) = 0 and u(x. 0) = sin x. (b) We have uy = f (y) where f is an arbitrary function of y.3 We have xux + (x + 1)yuy = u(1. 1) = e. 7.5 (a) The general solution to this equation is u(x) = C where C is an arbitrary constant. there is λ such that T X = =λ X tT and this leads to the two ODEs X = λX and T = λtT. u(x. Hence. 7.1 a = b = 0. y) = f (y) where f is an arbitrary function of y. That is.2 Subtituting into the diﬀerential equation we ﬁnd tX T − XT = 0 or T X = . This is true only when both sides are constant.6 (a) The general solution to this equation is u(x) = C1 x + C2 where C1 and C2 are arbitrary constants. a .235 Section 7 7. (b) The general solution is u(x. X tT The LHS is a function of x only whereas the RHS is a function of t only. 7.

t) = f (x + sin ut) . we must have 1 − tfv pu = 0. Thus (1 − tfv pu )ut = fv p.9 Let v = x + p(u)t. Hence. Using the chain rule we ﬁnd ut = fv · vt = fv · (p(u) + pu ut t). ANSWERS AND SOLUTIONS =2fv (v) + g(v) + 2xgv (v) =4fvv (v) + 4gv (v) + 4xgv (v) =fv (v) + xgv (v) =fvv (v) + xgvv (v) =2fvv (v) + gv (v) + 2xgvv (v) uxx − 4uxy + 4uyy =4fv (v) + 4gv (v) + 4xgv (v) −8fvv (v) − 4gv (v) − 8xgvv (v) +4fvv (v) + 4xgvv (v) = 0.7 Let v(x. 7. a contradiction. ut = Likewise. If ut = sin uux then p(u) = sin u so that the general solution is given by u(x. In this case. y) = y + 2x. 7.8 utt = c2 uxx . 1 − tfv pu It follows that ut = p(u)ux . 1 − tfv pu fv p . If 1 − tfv pu = 0 on any t−interval I then fv p = 0 on I which implies that fv = 0 or p = 0 on I. Then ux uxx uy uyy uxy Hence. ux = fv (1 + pu ux t) or ux = fv .236 7. But either condition will imply that tfv pu = 0 and this will imply that 1 = 1 − tfv pu = 0.

t) ≥ 0 for all x ∈ [0.237 where f is an arbitrary diﬀerentiable function in one variable.12 (a) This can be done by plugging in the equations. where F (x) = f (x)dx. t) = 18 e2x+3t . (d) We have sup{|un (x. t)|L − x=0 0 0 u2 (x. t) = u(L. t) = 0 and the fact that u2 (x. y) = xf (x − y) + g(x − y). x 7.15 We have ut utt ux uxx =cuv − cuw =c2 uvv − 2c2 uwv + c2 uww =uv + uw =uvv + 2uvw + uww Substituting we ﬁnd uvw = 0 and solving this equation we ﬁnd uv = f (v) and u(v. . n 2 (e) We have limt→∞ sup{|un (x. t)u(x. t) − 0 L u2 (x. we compute L L uxx (x. y) = xf (y − 2x) + g(y − 2x). 1 1 (c) We have sup{|un (x. the solution is unstable and thus the problem is ill-posed.10 u(x. t > 0} = limt→∞ Hence. w) = F (v) + G(w) where F (v) = f (v)dv. t)dx L =ux (L. t)u(L. 1 (c) u(x. 7. 3 2 (b) u(x. 7. 7.11 Using integration by parts. where f is an arbitrary function. (b) Plug in. t)dx =− 0 u2 (x. 7. t)u(x. y) = x 6y + F (x) + g(y). 0) − 1| : x ∈ R} = n sup{| sin nx| : x ∈ R} = n . t) − 1| : x ∈ R. en t n 2 = ∞. t) − ux (0. 7.13 (a) u(x. L]. t)u(0. t) − 1| : x ∈ R} = en t . t)dx ≤ 0 Note that we have used the boundary conditions u(0. y) = x3 + xy 2 + f (y).14 (b) u(x. t)dx = ux (x.

Also. t) = F (x + ct) + G(x − ct) where F and G are arbitrary diﬀerentiable functions. xy √ ) 2 y √ x − y 2√xy = x xy = u. 8.238 ANSWERS AND SOLUTIONS Finally. y) = y 2 . u(x. 8. 2 8. Then ux + 2uy − u = ex f (w) + 2ex fw (w) − 2ex fw (w) − ef (w) = 0. 8.2 Let w = 2x − y.4 We have −yux + xuy = −2xy sin (x2 + y 2 ) + 2xy sin (x2 + y 2 ) = 0. u(0. using the fact that v = x + ct and w = x − ct. 2 . 8.1 (a) Linear (b) Quasi-linear. Moreover.3 We have xux − yuy = x( xy + u(y.7 Plug u = av + w into the equation. Therefore.9 u(x. nonlinear. Using the linearity of L and the assumptions on v and w. t) = 1 (x + y) + f (x − y). we get d’Alembert’s solution to the one-dimensional wave equation: u(x. 1) = 1 (3−x2 ). √ 8.6 3a − 7b = 0. obtain Lu = L(av + w) = aLv + Lw = 0 + f = f for any constant a.8 us + c a2 +b2 = 0. 1 1 1 1 1 8. Moreover. 8.5 We have x ux + y uy = x (−x)+ y (1+y) = y . nonlinear (c) Nonlinear (d) Semi-linear. Section 8 8. y) = cos y 2 . u solves the inhomogeneous equation for any a.

9.11 We have ux = − 4e−4x f (2x − 3y) + 2e−4x f (2x − 3y) uy = − 3e−4x f (2x − 3y) Thus. 3ux + 2uy + 12u = − 12e−4x f (2x − 3y) + 6e−4x f (2x − 3y) −6e−4x f (2x − 3y) + 12e−4x f (2x − 3y) = 0. 9.2 u(x.13 u(x.15 vwv (v) = Aw(v). y) = f (ax − bt)e a . w).10 Using integration by parts we ﬁnd dI (t) = dt ∞ −∞ ∂u dx = ∂t ∞ xux dx −∞ ∞ = xf (x)|∞ − −∞ ∞ udx −∞ =− −∞ udx ≤ 0 Hence. 8. 8.3 u(x. t Section 9 9. 8.4 The change of coordinates v = x + t and w = x − t reduces the . y) = e c(ax+by) a2 +b2 f (bx − ay). y) = f (bx − ay). t) = sin (x − 3t). y) = x cos (2x − y) + f (y − 2x). 8.1 u(x.239 8. I(t) is decreasing. 9. 8.14 uw + λu = f (v + cw.12 u(x.

t) = 4 + x −t + g(x − t). 0) = f (x) − g(x). 0) − v(x.240 ANSWERS AND SOLUTIONS original equation to the equation uv = v+w whose solution is given by 2 2 2 (x+t)2 v2 wv u(v. Substituting these equations into the original equation we ﬁnd wt e−λt − λu + wx e−λt + λu = 0 or wt + wx = 0 9. t) is a solution to the equation follows from the principle of superposition. y) = f (Bx − Ay)e− A x . t) = e3t (x − t)2 + 9.11 u(x. C .13 (a) a = 1. w) = 4 + 2 + g(w) or u(x. t) = 9. 1−yh(x−y) 9. 0) = u(x. 9.7 Using the chain rule we ﬁnd wt = ut eλt + λueλt and wx = ux eλt . .12 u(x. y) = 1 (x + y) + f (x − y). But u(x. x. w(x. 9. 9. the given initial value problem has no solution. t)|} = max{|f (x) − g(x)|}. y) = h(x−y) . c = B.10 u(x. t) = e−3t . b = 0. 3 9.t x Thus.8 u(x. t) − v(x.6 u(x.9 (a) w(x. (c) From (b) we see that max{|u(x. x) = 1 2 so that 1 = x2 + g(0) or g(0) = x2 − 1 which is impossible since g(0) is a constant. t) = f (x − ct) − g(x − ct). t) = − sin 3t−2x 2 x c g t− e− c x λ 0 if x < ct if x > ct. Hence. Moreover. Hence.5 u(x. 1+(x−t)2 1 9 1 − 1t − 9. the problem is a well-posed problem. small changes in the initial data produces small changes in the solution. (b) w(x. 2 9. and d = −A (b) u(x.

3 The characteristics are parallel lines with common slope equals to 1 : x − y = k. 10. y 10.4 f y 1 .10 f (x2 + y 2 − u2 . 10. The char- . uy ) = 0 where f is an arbitrary diﬀerentiable function. 10. y x where f and g are arbitrary diﬀerentiable y 10. xy + u) = 0 where f is an arbitrary diﬀerentiable function. x2 + y 2 + u2 ) = 0 where f is an arbitrary diﬀerentiable function. y) = f (y − arctan x) for any diﬀerentiable function f. y x where f and g are arbitrary diﬀeren- x 10.8 f (x + y + z.6 f x . 9.1 The characteristics are hyperbolas: xy = k. 10.12 u(x. where f is an arbitrary diﬀerentiable function. 10.9 f (xy. xu = 0 or u = xn g n tiable functions. y) = f (x − y)e−x + x + y − 2. xy x 2 − arctan u = 0 where f is an arbitrary diﬀerentiable function. Section 10 10.13 u(x. y) = ex f (y − 2x).15 u(x. 10. y 10.2 The characteristics are circles centered at the origin: x2 + y 2 = k. 10. 10. y) = f (x − y)e−x . x4 − u4 − 2xyu2 ) = 0. y ln (y + u) − x) or u = −yg(y ln (y + u) − x) where f and g are arbitrary diﬀerentiable functions.11 f ( u .7 f x . u = 0 or u = xg x functions.241 9.14 u(x.5 f (y + u.

14 u = ex f (ye−x ) where f is an arbitrary diﬀerential function. 11. x+y x + y = 0. y where k is a constant.16 u = e− 2 f (ye−x ) where f is an arbitrary diﬀerentiable function of one variable. ANSWERS AND SOLUTIONS 10. y) = 1−xy . 11.1 u(x. 10. y) = (x + y)(x2 − y 2 ). Section 11 11.2 u(x.4 u(x. Solving this ODE 10.18 Solving dx = x by the separation of variables we ﬁnd x2 − y 2 = k.17 Solving dx = x by the separation of variables we ﬁnd x2 − y 2 = k. y) = ln x + 1 − y x . 2 x2 dy dx = x. dy 10. 11. . y where k is a constant.3 2xyu + x2 + y 2 − 2u + 2 = 0.15 The characteristics are solutions to the DE 2 we ﬁnd y = x + C.242 acteristics are shown below. dy 10.

y). 11. in the second case. y) = eyx .5 u(x. Since the solution is constant along any characteristic. Similarly.9 u(x. y) = y − sin−1 x.8 u(x. we’d need to ﬁnd a function f so that f (0) = h(x). 11. See ﬁgure below. y) as (x. 11. sec (x−ay)−y (x−1)2 2 11. 0). If h is not constant. x) = f (x − at). 0) along diﬀerent characteristics. y) = f (xe−y ). y) → (0. The characteristics are parobolas in the plane centered at the origin. 11. y) = h y − − (x − 1) ex−1 . Therefore.10 u(x. (ii) u(x. the method doesn’t work at that point. 0) is diﬀerent if we approach (0. if the solution is not exactly constant for all (x. 11.7 u(x.243 11. we cannot substitute x = 0 into yx−2 (the argument of the function f. it is not possible to satisfy this condition for all x ∈ R. (iv) All characteristics intersect at (0. y) = 1 . above) because x−2 is not deﬁned at 0. −2 . then the limit of u(x. (iii) In the ﬁrst case.11 (i) y = Cx2 .6 u(t. y) = f (x − uy).

dy 11. (b) We want 2 = u(x.13 (a) u = ex f (ye−x ) where f is an arbitrary diﬀerential function. u(x. y) = f (x2 − y 2 ) where f is 2 an arbitrary diﬀerentiable function. 11. Solving this equay tion we ﬁnd x2 − y 2 = C. This equation is impossible so this Cauchy problem has no solutions. The solution in the region y 2 − x2 < 0 can be . Since u(0.244 ANSWERS AND SOLUTIONS 11. y) = −1 + 2e 2 e− x2 (4x−y)2 2 . (d) This solution is only deﬁned in the region covered by characteristics that cross the y axis: y 2 − x2 > 0.16 (a) The characteristics satisfy the ODE dx = x . 2 −y 2 Hence. y) = ex f (ye−x ) where f is an arbitrary function satisfying f (1) = 1.14 u(x. 3x) = ex f (3ex e−x ) = ex f (3). the characteristics are hyperbolas. there are inﬁnitely many solutions to this Cauchy problem. 11. u(x. y) = ey cos (x − y). Thus. 11. y) = e−y we ﬁnd f (y) = ey .15 The Cauchy problem has no solutions. In this case.12 u(x. (b) (c) The general solution to the PDE is u(x. (c) We want ex = ex f (ex e−x ) =⇒ f (1) = 1. y) = ex . namely.

Thus. t) utt (x.2 (a) Ellitpic (b) Hyperbolic (c)Hyperbolic. 0) ut (x.17 (a) Solving the ODE dx = y we ﬁnd the characteristics ye−x = C. • The PDE is of elliptic type if 4y 2 (x2 + x + 1) < 0 which can not happen. cos x cos t. t) uxx (x. u(x.245 any function of the form u(x. Since x2 + x + 1 > 0 for all x ∈ R. − cos x sin t. t) u(x. 0) = x. t) = = = = − cos x sin t = utt (x. • The PDE is of parabolic type if 4y 2 (x2 + x + 1) = 0. = = = = − sin x sin t. cos x cos 0 = cos x. 0) = 1 then we choose f to be any arbitrary diﬀerentiable function satisfying f (0) = 1. − sin x sin 0 = 0. t) ut (x. If u(x. 12. 12. Section 12 12. Graphically. 0) ux (0. − cos x sin t. 12. this is x−axis. Hence.5 (a) Quasi-linear (b) Semi-linear (c) Linear (d) Nonlinear. This is true for all y = 0 and x2 + x + 1 > 0. dy 11. . t) Thus. Graphically. t). (b) The line y = 0 is a characteristic so that u has to be constant there. 12. uxx (x. there is no solution satisfying the condition u(x. y) = f (x2 − y 2 ).3 • The PDE is of hyperbolic type if 4y 2 (x2 + x + 1) > 0. this is the inside of the parabola x2 + x + 1 that opens up.4 We have ux (x. y) = f (ye−x ).1 (a) Hyperbolic (b) Parabolic (c) Elliptic. we must have y = 0. − sin 0 sin t = 0.

12. kL > 0. y) = f (x)g(y) into the left side of the equation to obtain f (x)g(y)(f (x)g(y))xy = f (x)g(y)(f (x)g (y)).246 12. which means f (x)g(y) is a solution. So the sides are equal. Therefore. 12. substitute the same thing into the right side to obtain (f (x)g(y))x (f (x)g(y))y = f (x)g(y)f (x)g (y) = f (x)g(y)f (x)g (y). because k0 .9 We have uxx = −n2 sin nx sinh ny and uyy = n2 sin nx sinh ny Hence. Similar argument holds for the second part of the problem. where F (x) = f (x)dx. ∆u = 0. the right-hand side is nonpositive and the left-hand side is nonnegative.8 Substitute u(x. we obtain L L λ 0 u2 (x)dx = 0 uuxx (x)dx L =[u(L)ux (L) − u(0)ux (0)] − 0 L u2 (x)dx x u2 (x)dx x = − kL u2 (L) + k0 u( 0) + x x 0 For λ > 0.10 u(x. which is the trivial solution. y) = x2 y 2 4 + F (x) + G(y). . Now.6 We have ANSWERS AND SOLUTIONS 2x x2 + y 2 2y 2 − 2x2 uxx = 2 (x + y 2 )2 2y uy = 2 x + y2 2x2 − 2y 2 uyy = 2 (x + y 2 )2 ux = Plugging these expressions into the equation we ﬁnd uxx + uyy = 0. and there can be no solution other than u ≡ 0. 12. both sides must be zero. 12.7 Multiplying the equation by u and integrating.

B = −2 cos x. (b) 1 + 4x2 y = 0. So this equation is hyperbolic everywhere in R2 . 12. 12. t) = (f (x + ct) + f (x − ct)) + [g(x + ct) − g(x − ct)) 2 2c 1 1 uxx (x. It is hyperbolic if (x − 1)2 + (y + 1)2 > 4 and elliptic if (x − 1)2 + (y + 1)2 < 4. C = 7 so B 2 −4AC = 16−56 = −40 < 0. y) = f (y − 3x) + g(x + y) = 10x2 +y 2 −7xy+6 . C = − sin2 x so B 2 − 4AC = 4 cos2 x + 4 sin2 x = 4 > 0.13 (a) 1 + 4x2 y > 0. 12.247 12. 1 1 u(x. y) = f (y − 3x) + g(x + y). So this equation is elliptic everywhere in R2 .12 Using the chain rule we ﬁnd 1 1 ut (x.14 u(x. 0) = (f (x) + f (x)) + 2 2c and ut (x. (c) 1 + 4x2 y = 0.11 (a) We have A = 2. 0) = g(x). 12. (b) We have A = 1.15 u(x. (c) We have A = y. The equation is parabolic if (x − 1)2 + (y + 1)2 = 4. B = 2(x − 1). t) = (cf (x + ct) − cf (x − ct)) + [g(x + ct)(c) − g(x − ct)(−c)) 2 2c c 1 = (f (x + ct) − f (x − ct)) + (g(x + ct) + g(x − ct)) 2 2 c2 c utt = (f (x + ct) + f (x − ct)) + (g (x + ct) − g (c − xt)) 2 2 1 1 ux (x. C = −(y + 2) so B 2 − 4AC = 4(x − 1)2 + 4y(y + 2) = 4[(x − 1)+ (y + 1)2 − 4]. 6 x g(s)ds = f (x) x Section 13 . B = −4. t) = (f (x + ct) + f (x − ct)) + [g (x + ct) − g (x − ct)) 2 2c By substitutition we see that c2 uxx = utt . Moreover.

t)ux (L.7 1 [sin (2π 8π + 4πt) − sin (2π − 4πt)]. t) − c2 0 2 ut uxx dx =c ut (L. 13. t) = 1 + 13. 13. t)ux (0.3 u(x.10 v(r) = A cos nr + B sin nr. t)) . t) + 0 ut (utt − c2 uxx )dx =c (ut (L.2 Indeed we have c2 uxx (x.9 Just plug the translated/diﬀerentiated/dialated solution into the wave equation and check that it is a solution. 1 13. 13. t)ux (0.11 u(x. t) = 1 2 1 1+(x+t)2 + 1 1+(x−t)2 . t) − c ut (0. 13. t) = 2 (cos (x − 3t) + cos (x + 3t)). 13. t) = αv(x. c 13.   1 if x − 5t < 0 and x + 5t < 0 1 if x − 5t < 0 and x + 5t > 0 u(x.8 u(x. 2 2 t 1 13. t). t). 13. t)ux (0. Then we have c2 zxx =c2 αvxx + c2 βwxx =αvtt + βvtt =ztt . t) = 2 [e−(x+ct) + e−(x−ct) ] + 2 + 1 4c cos (2x) sin (2ct).6 u(x. t) =  2 0 if x − 5t > 0. t)ux (L. 1 13.248 ANSWERS AND SOLUTIONS 13. t) + βw(x.12 (a) We have dE (t) = dt = 0 L 2 2 L L (ut utt + 0 L 0 c2 ux uxt )dx L ut utt dx + c2 ut (L.1 Let z(x. t) = 2 [ex−ct + ex+ct + 1 (cos (x − ct) − cos (x + ct))]. t) − ut (0.4 u(x. t)ux (L.5 u(x. t) = 0 = utt (x. t) = 0. t) − c2 ut (0.

Thus. The latter case can occur only if ut (x. t)ux (L. Then we have kzxx =kαuxx + kβvxx =αut + βvt =zt . t) is identically zero. t) = c2 R (x − ct). t) = 0 = ut (x.4 Let u be the solution to (14. utt = c2 uxx . 13. we ﬁnd dE (t) = −d dt L (ut )2 dx. t) = x2 + 4t2 − 4 sin 2x sin 4t. t) = u(L. 0 The right-hand side is nonpositive. Likewise.2 Indeed we have kuxx (x.15 u(x.1 Let z(x. From (a) we have dE (t) = c2 (ut (L.13 Using the previous exercise. t) = 0.14 (a) By the chain rule we have ut (x. 14. Let . t)) = 0. that is ux (0. 14. dt (c) Assuming free ends boundary conditions. t) − ut (0. t) = αu(x. t) = 0. so the energy either decreases or is constant.1) that satisﬁes u(0. t) = 0. t). t) = −cR (x − ct) and utt (x. ux (x. t) = R (x − ct) and uxx = R (x − ct). we ﬁnd dE (t) = 0. dt 13. t)ux (0. 2 1 13. t) = T0 + TL −T0 x. we have ut (0. L 14. Section 14 14. which means that the string is at rest. t). t) = ux (L.3 u(x. (b) Since the ends are ﬁxed.249 since utt − c2 uxx = 0.— (b) We have L L (ut )2 dx = 0 0 c2 [R (x − ct)]2 dx = 2 L 0 c2 (ux )2 dx. t) = ut (L. t) + βv(x.

Since L > 0. 14.1) that satisﬁes w(0. the function u(x. the condition u(L.6 Substituting u(x.1) we obtain k T X = .7 (a) Letting α = λ > 0 we √ obtain the √ ODE X − αX = 0 whose general k αx − αx solution is given by X(x) = Ae + Be for some constants A and B. 14. This gives the two ordinary diﬀerential equations X − λ X = 0 and T − λT = 0. 14. (b) The condition u(0. t) is a solution to (14.1. t) = TL . Since c2 = 0 we must have sin βL = 0. A(e λL − e− λL ) = 0. the number on the left is greater than 1 whereas the number on the right is less than 1. That is.8 (a)Now. . t) = T0 and u(L. t) = T0 + TLL 0 x. √ Likewise.250 ANSWERS AND SOLUTIONS w(x. Hence. t) = T0 −T and w(L. Hence. t) = u(x. write σ = −λ > 0. Then we obtain the equation X + σ X = 0 k whose general solution is given by X(x) = c1 cos βx + c2 sin βx where β = σ = −λ . a contradiction. t) is the trivial solution which contradicts the assumption that u is non-trivial. k 14. X T Since X only depends on x and T only depends on t. k k (b) Using X(0) = 0 we obtain c1 = 0. √ Hence. t) = TL . t) = 0. Hence. From Exercise 14. t) + w(x. we √ (d) Using (b) and (c) we obtain e λL = e− λL . (c) If A = 0 then B = 0 and u(x. t) = 0 implies Ae λL + Be− λL = 0. we must have A = B = 0 which leads to the trivial solution. we must have that there is a constant λ such that k X = λ and X T T = λ.1) that satisﬁes u(0. we must have λ < 0. t) be the time independent solution to (14. √ must have A = 0.5 u(x. t) = 0 implies that X(0) = 0 which in turn implies √ √ A + B = 0. w(x. t) = X(x)T (t) into (14.

t) = 0. L (b) v(x) = T.10 (i) u(0. 14. . (c) v(x) = αx + T. L < x < 2L. u(x) = u1 (x) 0 < x < L u2 (x) L < x < 2 The steady temperature satisﬁes the one-dimensional Laplace equation. so we have u1 = 0.11 Solving this problem we ﬁnd u(x. t) is also a solution to (14. 14. t) = 100 for t > 0.9 For each integer n ≥ 0 we have un (x. kn2 π 2 n 14. E (t) = −2πe2 t < 0 for all t > 0. 0 < x < L. t) = 0 since un (0. Thus. u2 = 0. (ii) ut (0.1). Moreover. L < x < 2L. and denote by u1 (x) and u2 (x) the temperature distributions in the respective intervals. u(0.1). t) = 0 and u(a. t) = un (L. 14. 14. Solving the equations. u(x.251 Thus. t) = Tc(0) T (0)e L2 t sin nπ x L is a solution to (14. t) = u(L. σ = kn2 π 2 . L2 where n is an integer. t) = ut (a.13 v(x) = x + 2. By superposition.12 E(t) = L 0 f (x)dx + (1 + 4L)t. we ﬁnd u1 (x) = C1 x + C2 . t) = e−t sin x. We have π π E(t) = 0 [e−2t sin2 x + e−2t cos2 x]dx = 0 e−2t dt = πe−2t .15 (a) We ﬁrst consider the intervals 0 < x < L and L < x < 2L independently.14 (a) v(x) = T x. 0 < x < L. 2 (d) v(x) = − x + T2 −T1 + 2 L L 2 x + T1 . Thus. 14. t) = 0 for t > 0. u2 (x) = C3 x + C4 . 14.

C3 = · . T1 = 0.16 (a) E(t) = 0 cρu(x. u2 (x) = T2 + C3 (x − 2L). L (b) If L = 1. 0 . 3 14. from the continuity of the temperature. 3 3 3 100 3 The temperature is a piecewise linear function which has slope x < 1 and slope 200 for 1 < x < 2. From the last condition. t)dx. t)|L = 0. u2 (2L) = T2 (boundary conditions) u1 (L) = u2 (L) (continuity of the temperature) k1 u1 (L) = k2 u2 (L) (continuity of the heat ﬂux) From the ﬁrst two conditions. k1 = 2. k2 Finally. (b) We integrate the equation in x from 0 to L : L L L for 0 < cρut (x. T1 + C1 L = T2 − C3 L =⇒ C1 + C3 = and we obtain C1 = Thus. u2 (x) = T2 + (T2 − T1 ) .252 ANSWERS AND SOLUTIONS We have the following conditions to determine the constants: u1 (0) = T1 . k1 C1 = k2 C3 =⇒ C3 = k1 C1 . u1 (x) = T1 + (T2 − T1 ) k1 x − 2L k2 x . k1 + k2 L k1 + k2 L k2 T2 − T1 k1 T2 − T1 · . u2 (x) = x− . T2 = 100. u1 (x) = T1 + C1 x. k2 = 1 we ﬁnd u1 (x) = 100 200 100 x. k1 + k2 L k1 + k2 L T2 − T1 . t)dx = 0 0 Kuxx dx = Kux (x.

We have dE = dt L L ut (x. 6 0 0 Performing the integration and then solving for C2 we ﬁnd C2 = 1 L L f (x)dx + 0 L3 L −β . This means L L x3 − + βx + C2 dx = f (x)dx = E(0). 2 (b) The steady solution (equilibrium) is possible if the right-hand side vanishes: L2 =0 (7 − β) + 2 Solving this equation for β we ﬁnd β = 7 + L . 0 Thus.17 (a) The total thermal energy is L E(t) = 0 u(x. t) = 0. 24 2 Therefore. The left-hand side can also be written as d dt L cρu(x. 14. The steady solution should also have the same value of the total energy as the initial condition. t) = ux (L. 2 (c) By integrating the equation uxx + x = 0 we ﬁnd the steady solution u(x) = − x3 + C1 x + C2 6 2 From the condition ux (0) = β we ﬁnd C1 = β. t)dx = E (t). 24 2 6 . we have shown that E (t) = 0 so that E(t) is constant.253 since ux (0. t)dx = ux |L + 0 0 0 xdx = (7 − β) + L2 . the steady-state solution is u(x) = 1 L L f (x)dx + 0 L3 L x3 − β + βx − . t)dx.

254 ANSWERS AND SOLUTIONS Section 15 15. 2 (b) For n = m we have L cos −L mπ nπ 1 x sin x dx = L L 2 (m + n)π (m − n)π x − sin x L L −L (m + n)π L 1 cos x =− 2 (m + n)π L sin + L cos (m − n)π (m − n)π x L L −L L dx =0 where we used the trigonometric identiy 1 cos a sin b = [sin (a + b) − sin (a − b)].3 (a) L (b) L (c) 0. .2 (a) For n = m we have L sin −L nπ 1 L (m + n)π mπ x sin x dx = − cos x − cos L L 2 −L L 1 L (m + n)π =− sin x 2 (m + n)π L L sin − (m − n)π =0 (m − n)π x L L −L (m − n)π x L dx where we used the trigonometric identiy 1 sin a sin b = [− cos (a + b) + cos (a − b)]. 15. (b) We have (c1 f +c2 g)(x+T ) = c1 f (x+T )+c2 g(x+T ) = c1 f (x)+c2 g(x) = (c1 f + c2 g)(x). 2 15.1 (a) We have (f g)(x + T ) = f (x + T )g(x + T ) = f (x)g(x) = (f g)(x).

and in particular we can write g(a) = g(0) for all a ∈ R which gives the desired result. the function is not piecewise continuous in [−1. . 15.6 f (x) = 15. 1]. Using the fundamental theorem of calculus. nx 2 cos nπ 2 − (−1)n sin nπ x 2 .8 Since the sided limits at the point of discontinuity x = 0 do not exist. 15.9 Deﬁne the function L+a g(a) = −L+a f (x)dx.4 1 π 1 an = π a0 = =− bn = 1 π −π π π f (x)dx = 0 −π π f (x) cos nxdx −π 0 π cos nxdx + 0 cos nxdx = 0 f (x) sin nxdx −π 0 π =− −π sin nxdx + 0 sin nxdx 2 = [1 − (−1)n ] n 15.5 f (x) = − 1 + 6 15. ∞ 4 n=1 (nπ)2 [1 − (−1)n ] cos .7 f (x) = ∞ 2 n=1 nπ ∞ 4 n n=1 (nπ)2 (−1) cos (nx).255 15. g is a constant function. we have L+a dg d = f (x)dx da da −L+a =f (L + a) − f (−L + a) = f (−L + a + 2L) − f (−L + a) =f (−L + a) − f (−L + a) = 0 Hence.

because this function and its derivative are piecewise continuous. n ∈ N. 15. 1 (c) a0 = 1. b1 = 1.11 (a) a0 = 2.10 (i) f (x) = 10 + ∞ [− nπ sin 2nπ cos 2nπx +(− nπ − cos 2nπ + 1 ) sin n=1 3 3 3 3 (ii) Using the theorem discussed in class. (iii) 2nπx 3 .13 an = 0 for all n ∈ N. . 4 Section 16 16. At any point of discontinuity. an = 0. (b) a0 = 4.14 f (0− )+f (0+ ) 2 = 3 2 −π+π 2 = 0. an = 0. 2L (d) a0 = an = 0.2 . ∞ sin (2n−1)x . the Fourier series will converge to the average of the left and right limits.1 f (x) = 0. 16. 15. and bn = 0. an = bn = 0 for n ∈ N. the Fourier series will converge to the function at each point of continuity.15 (a) f (x) = (b) ∞ n=1 (−1)n+1 2n−1 + 2 π = π.12 −1 15.256 ANSWERS AND SOLUTIONS 1 1 15. bn = πn [1 − (−1)n ]. n=1 2n−1 15. bn = πn (−1)n+1 . n ∈ N. 15.

4 .3 16.257 16.

2 (ii) By the theorem discussed in class.7 f (x) = 16. 1+(−1)n n2 −1 n sin nx. respectively a0 nπx + an cos . cos (nπx).5 f (x) = 16. 2] are. − 1] cos nx. ∞ 2 n=1 nπ [1 2 pi ∞ n=1 − (−1)n ] sin nx.8 f (x) = π 4 π 2 + + ∞ 2 n=1 πn2 [2 cos (nπ/2) ∞ 2 n n=1 n2 π [(−1) − 1 − (−1)n ] cos nx. 1 16.9 f (x) = 2 (e2 − 1) + ∞ 4 n=1 4+n2 π 2 16.10 (i) The formulas for the cosine and sine series of a function f (x) on an interval [0. bn = 2 2 x sin 0 nπx dx. the Fourier sine series will converge pointwise to f on the open . since the function is continuously differentiable .6 f (x) = 16.258 ANSWERS AND SOLUTIONS 16. an = 2 2 n=1 ∞ ∞ 2 x cos 0 nπx dx 2 bn sin n=1 nπx .

However. On the left. the average of the left and right limits of the odd extension there. . and so its cosine series will be pointwise convergent everywhere. 2]. (iv) The key point is that the even extension is continuous on [−2. which is zero. (iii) By the theorem discussed in class. and that of odd extension is shown on the right. the limit of the sine series. 2). The graph of the even extension is shown on the left. so there the sine series will converge to the average of its left and right limit. the Fourier cosine series will converge pointwise to f on the open interval (0. the limit of the cosine series. 2). On the right. its odd extension has a jump discontinuity at x = ±2. which converges pointwise to the even extension on the entire closed interval.259 interval (0. since the function is continuously diﬀerentiable . where it converges to 0. which converges pointwise to the odd extension everywhere except at x = ±2.

nπ 2 x dx = [1 − (−1)n ]. n ∈ N. 2 16.13 By deﬁnition of Fourier sine coeﬃcients. .12 (a) a0 = 10 and a1 = 1. bn = The symmetry around x = f 2 L L f (x) sin 0 nπ x dx L L 2 can be written as L +y 2 =f L −y 2 for all y ∈ R.11 (a) If f (x) = sin (b) If f (x) = 1 then 2 bn = L (c) If f (x) = cos π x L 2π x L ANSWERS AND SOLUTIONS then bn = 0 of n = 2 and b2 = 1. 2 (c) a0 = 1 and an = πn sin πn . (n − 1)π L 0 16.260 16. To use this symmetry it is convenient to make the change of variable x = L + y in the above integral to obtain 2 L 2 bn = −L 2 f L nπ + y sin 2 L L +y 2 dx. 2L (b) a0 = L and an = (πn)2 [(−1)n − 1]. L nπ L sin 0 then 2 L L b1 = and for n = 1 we have bn = cos 0 π π x sin x dx = 0 L L π π 2 L cos x sin x dx L 0 L L 12 L πx πx = sin (1 + n) − sin (1 − n) dx 2L 0 L L = 1 L πx L πx − cos (1 + n) + cos (1 − n) L (1 + n)π L 1−n L 2n = 2 [1 + (−1)n ]. and an = 0 for n = 1. n ∈ N.

2 2 16. Since f L + y is odd in y and for n even cos nπ L + y = cos nπy is 2 L 2 L even in y.261 Since f L + y is even in y and for n even sin nπ L + y = sin nπy is 2 L 2 L odd in y. 2.14 By deﬁnition of Fourier cosine coeﬃcients. the integrand of the above integral is odd in y for n even. 1. To use this symmetry it is convenient to make the change of variable x = L + y in the above integral to obtain 2 L 2 an = −L 2 f nπ L + y cos 2 L L +y 2 dx. 2 2 16. .15 sin 16. Since the intergral is from − L to L we must have an = 0 for all n = 0. Since the intergral is from − L to L we must have bn = 0 for all n ∈ N. an = 2 L L f (x) cos 0 L 2 nπ x dx L The anti-symmetry around x = f can be written as = −f L +y 2 L −y 2 for all y ∈ R.16 (a) nπx L = 2 π − 2 π ∞ 1+(−1)n n=2 n2 −1 cos nπx L . the integrand of the above integral is odd in y for n even. · · · .

. X(x) Y (y) The left hand side is a function of x while the right hand side is a function of y. X(x) Y (y) where δ is a constant. That is. dividing for X(x)Y (y) and subtract both sides for X (x) . − X (x) Y (y) = + λ = δ. This says that they must equal to a constant.1 We look for a solution of the form u(x. we ﬁnd: X(x) − X (x) Y (y) = + λ. (e) ∞ 2 nπ 3 − sin f (x) = + 2 n=1 nπ 2 cos nπx . Substituting in the given equation.262 (b) a0 = 2 0 f (x)dx = 3. Assuming X(x)Y (y) is nonzero. 2 Section 17 17. nπ 2 (d) bn = 0 since this is an even function. we obtain X Y + XY + λXY = 0. 2 (c) We have 2 an = 2 2 2 ANSWERS AND SOLUTIONS f (x) cos 0 1 nπx dx 2 2 nπx nπx cos 2 cos = dx + dx 2 2 0 1 1 2 2 nπx 2 nπx = sin +2 sin nπ 2 nπ 2 0 1 2 nπ =− sin . This results in the following two ODEs X + δX = 0 and Y + (λ − δ)Y = 0. y) = X(x)Y (y).

Substituting into the heat equation we obtain X T = . we must have that there is a constant λ such that X X = λ and T kT = λ.2 Let’s assume that the solution can be written in the form u(x. t) = X(x)T (t). 17.263 • If δ > 0 and λ − δ > 0 then X(x) =A cos δx + B sin δx Y (y) =C cos (λ − δ)y + D sin (λ − δ)y • If δ > 0 and λ − δ < 0 then X(x) =A cos δx + B sin δx √ √ − −(λ−δ)y Y (y) =Ce + De −(λ−δ)y • If δ = λ > 0 then X(x) =A cos δx + B sin δx Y (y) =Cy + D • If δ = λ < 0 then X(x) =Ae− −δx + Be Y (y) =Cy + D • If δ < 0 and λ − δ > 0 then X(x) =Ae− −δx + Be −δx Y (y) =C cos (λ − δ)y + D sin (λ − δ)y • If δ < 0 and λ − δ < 0 then X(x) =Ae− √ −δx √ √ √ √ √ −δx Y (y) =Ce− √ + Be −δx √ (λ−δ)y (λ−δ)y + De . . X kT Since X only depends on x and T only depends on t.

T = λT. t) = eλ(x−t) .6 u(x. Assuming X(x)T (t) is nonzero. 17. 17. T = kλT.9 u(x. 17. Substituting in the wave equation. Solving these equations we ﬁnd X(x) = ax + b and T (t) = c. Case 3: λ < 0 √ √ In this case. Case 1: λ = 0 In this case.7 5X − 7X − λX = 0 and 3Y − λY = 0. 17. dividing for X(x)T (t) we ﬁnd: X (x) T (t) = .5 X − λX = 0. y) = X(x)T (t). we consider the three cases of the sign of λ. X(1) = 0.4 X = (2 + λ)X. Case 2: λ > 0 √ √ In this case. Next.8 u(x. X(x) = A cos −λx + B sin −λx and and T (t) = Cekλt .264 ANSWERS AND SOLUTIONS This gives the two ordinary diﬀerential equations X − λX = 0 and T − kλT = 0. X(0) = 0. X (0) = 0 = X (L). 17. 17. X(x) = Ae λx + Be− λx and T (t) = Cekλt .3 r2 R (r) + rR (r) − λR(r) = 0 and Θ (θ) + λΘ(θ) = 0. we obtain X (x)T (t) − X(x)T (t) = 0. X(x) T (t) y . y) = Ceλx− λ . X = 0 and T = 0. 17. 17. y) = Ceλx y λ .10 We look for a solution of the form u(x.

265 The left hand side is a function of x while the right hand side is a function of t. √ √ √ √ u(x. 17. t) = R(r)T (t). C. X (x) T (t) = =λ X(x) T (t) where λ is a constant. and D are arbitrary constants.11 (a) u(r. t) = k1 xt + k2 x + k3 t + k4 . B. B. In this case. . C. This results in the following two ODEs X − λX = 0 and T − λT = 0. u(x. and D are constants. t) = k1 eλ(x+t) + k2 eλ(x−t) + k3 e−λ(x+t) + k4 e−λ(x−t) . The solutions of these equations depend on the sign of λ. and C are arbitrary constants. In this case. • If λ < 0 then √ √ X(x) =A cos −λx + B sin −λx √ √ T (t) =A cos −λt + B sin −λt where A. t) =k1 cos −λx cos −λt + k2 cos −λx sin −λt √ √ √ √ +k3 sin −λx cos −λt + k4 sin −λx sin −λt. 1 (rR r ) = −λR. • If λ > 0 then the solutions are given √ X(x) =Ae T (t) =Ce λx λt + Be− + De− √ λx √ √ λt where A. That is. In this case. B. T (t) = −kλT. u(x. • If λ = 0 then X(x) =Ax + B T (t) =Ct + D where A. This says that they must equal to a constant.

7 u(x. nπ 18. 9. 5. T = −kλT.1 u(x. t) = X(x)T (t). X = −λX. · · · Cn =  6 n is odd. Y − Y + Y = λY. t) = 6 sin 18. 17. 6. 8. 12.6 u(x. t) = 6 + 4 cos 18. 2L . T = −λT.12 u(x. · · · Cn =  nπ 0 n is even 3π x L 18. t) = 18. n2 π 2 + Cn cos nπ x e− L2 t where L  2  − nπ n = 1. (e) u(x. kX − aX = −λX. R = −λR.266 ANSWERS AND SOLUTIONS (b) u(x. (c) u(x. 64π 2 t L2 9π 2 8π x L e− . t) = −3 cos 18.9 u(x. t) = X(x)Y (y). 10. ∞ n=1 Cn sin nπ x L e− n2 π 2 t L2 where  4  − nπ n = 2.2 u(x. 7. t) = 8d π3 2 π π x 2 e− π2 k t 4 + 3 sin sin 5π x 2 e− 25π 2 k t 4 . ∞ 1 n=1 (2n−1)3 (2n−1)π x L e− k(2n−1)2 π 2 t L2 − 4 π ∞ 1 n=1 (4n2 −1) cos 2nπ x L e−k 4n2 π 2 t L2 . T + αT = −kλT. Y = −λY.13 X = λX. 11. · · · 2 n = 3. 17. t) = 1 2 9π x L ∞ n=1 e −81π 2 t L2 . t) = e− L2 t . · · · 0 n = 4. X = −λX. t) = X(x)T (t). (d) u(x.4 u(x.5 u(x. t) = X(x)T (t). . t) = 18. ∞ an cos n=0 (2n+1)2 π 2 (2n + 1)π x e− 4L2 t . y) = Ceλ(x+y) . Section 18 18.8 u(x.3 u(x. t) = sin 18.

e − 1+ n 2 π2 L2 t → 0 for each n ∈ N. (c) Since w(x. t)dx + 0 0 0 =2 = 2w(x. t)[wxx (x. t) = 0 and ux (1. This gives the two ordinary diﬀerential equations X − λX = 0 and T − kλT = 0. E(t) = 0 for all t ≥ 0. u(x. . (b) Let’s assume that the solution can be written in the form u(x. u(x. we must have that there is a constant λ such that X X = λ and T kT = λ. t) → 0. t)wx (x. 0) = 0. and 0 ≤ E(t) ≤ E(0) for all t > 0. Hence. t)dx w(x. Substituting into the heat equation we obtain X T = . t) → 0. t)dx + 0 1 2 wx (x. Hence. ∞ an cos (nπ − x e L 1+ n 2 π2 L2 t . As t → ∞. t)]dx 0 1 1 2 wx (x. we must have E(0) = 0. t). Hence. t)wt (x. t)dx ≤ 0 Hence. This implies that w(x. t)|1 − 2 0 1 w2 (x. t)dx =−2 w2 (x.267 As t → ∞. t) = 0.10 u(x. t) = u2 (x. t) − w(x. t) = 0 for all t > 0 and all 0 < x < 1.12 (a) u(0. E is decreasing. This means that the given problem has a unique solution. e− 18. 18. t) = X(x)T (t). Therefore u1 (x. 18.11 (b) We have 1 E (t) =2 0 1 w(x. X kT Since X only depends on x and T only depends on t. t) = n=0 (2n+1)2 π 2 t 4L2 → 0 for each n ∈ N.

Next. As far as the boundary conditions. λ = − n − 1 π 2 . Substituting into the heat equation we obtain T X = . X −λX = 0. X = 0. 2 18. Case 1: λ = 0 In this case. t) = 0 = X(0)T (t) =⇒ X(0) = 0 and u(L.13 (a) Let’s assume that the solution can be written in the form u(x. Thus. Note that T is not the zero function for otherwise u ≡ 0 and this contradicts our assumption that u is the non-trivial solution. X kT Since the LHS only depends on x and the RHS only depends on t. Now. Hence. X ≡ 0 and . t) = 0 = X (1)T (t) =⇒ X (1) = 0. then there must be a constant λ such that X X = λ and T kT = λ. t) = 0 = X(L)T (t) =⇒ X(L) = 0. Solving this equation we ﬁnd X(x) = ax + b. X (1) = 0 implies cos −λ = 0 or −λ = 2 1 n − 2 π. t) = X(x)T (t). Since X(0) = 0 we ﬁnd b = 0. Note that T is not the zero function for otherwise u ≡ 0 and this contradicts our assumption that u is the non-trivial solution. t) = 0 = X(0)T (t) =⇒ X(0) = 0 and ux (1. we have ANSWERS AND SOLUTIONS u(0. we have u(0. Hence. Since X(L) = 0 we ﬁnd a = 0. √ √ Moreover X(0) = 0.√ √ √ (c) We have X = −λ cos −λx and X = λ sin −λx.268 As far as the boundary conditions. n ∈ N. This gives the two ordinary diﬀerential equations X − λX = 0 and T − kλT = 0. we consider the three cases of the sign of λ.

Again. L . t) ≡ 0. satisfy ut = kuxx and the boundary conditions u(0. t) = n=1 Cn sin n2 π 2 nπ x e− L2 kt . The√ implies A = 0. Solving for λ 2 2 we ﬁnd λ = − nLπ . in order for these solutions to satisfy the initial value condition u(x. Case 3: λ < 0 √ √ condition X(0) = 0 In this case. solving the equation T − λkT = 0 by the method of separation of variables we obtain Tn (t) = Bn e− Hence. That is. 0) = 6 sin 9πx . the conditions X(0) = X(L) = 0 imply A = B = 0 and hence the solution is the trivial solution. we invoke the superposition principle of linear PDE to write L ∞ u(x. The condition X(L) = 0 implies B sin −λL = 0. t) = Cn sin n2 π 2 nπ x e− L2 kt . n ∈ N L n2 π 2 kt L2 nπ x. We must have B = 0 otherwise√ X(x) = 0 and this leads to the trivial solution.269 u(x. Case 2: λ > 0 √ √ In this case. u is the trivial solution. n ∈ N. X(x) = A cos −λx + B sin −λx. L (1) To determine the unknown constants Cn we use the initial condition u(x. we obtain inﬁnitely many solutions given by 2 Xn (x) = An sin Now. Now. Thus. t) = u(L. we obtain sin −λL = 0 or −λL = nπ where n ∈ N. L . 0) = 6 sin 9πx in (1) to obtain L 6 sin 9πx L ∞ = n=1 Cn sin nπ x . Since √ B = 0. the functions un (x. t) = 0. n ∈ N. X(x) = Ae λx + Be− λx .

πx L π 2 kt π x e− L2 − sin L 9π 2 kt 3π x e− L2 L 9πx − 81π2 kt e L2 . y) = 2xy then uxx = uyy = 0 so that ∆u = 0. y) = ∞ n=1 f2 (y) sin 0 nπ y dy b nπ (y a .15 u(x. Hence.4 If u(x.3 u(x.5 u(x.2 u(x.270 ANSWERS AND SOLUTIONS By equating coeﬃcients we ﬁnd C9 = 6 and Cn = 0 if n = 9. y) = x2 − y 2 then uxx = 2 and uyy = −2 so that ∆u = 0. y) = n=1 ∞ [An cosh nπ nπ nπ y + Bn sinh y ] sin x. t) = 5. If u(x. L e− pi2 kt L2 + 4 cos 5πx L e− 25pi2 kt L2 . 19. a a sin πx sinh πy.14 u(x. 18. t)6 sin (b) Similar to (a). L L L −1 where 2 An = L L 0 nπ (f1 (x) + f2 (x)) sin xdx L cosh nπH 2L . t) = 6 sin xe−8t . the solution to the problem is given by u(x. y) = ∞ n=1 Bn sin b nπ y b sinh nπ x b where sinh nπ a b −1 2 Bn = b 19. Bn sin nπ x sinh a a − b) where 2 Bn = a 19. 19. we ﬁnd u(x. t) = 3 sin 18. y) = 2xy + g1 (x) sin 0 3 sinh π nπ nπ x dx [sinh − b ]−1 . Section 19 19.1 u(x. t) = cos (b) u(x.

Multiply this equation by i we ﬁnd ∞ −iuy + vy = n=0 nan (x + iy)n−1 . Hence.7 Polar and Cartesian coordinates are related by the expressions x = 1 y r cos θ and y = r sin θ where r = (x2 + y 2 ) 2 and tan θ = x . (b) We have uxx = (vy )x = (vx )y = −uyy so that ∆u = 0. Using the chain . Likewise. Similar argument for ∆v = 0.6 (a) Diﬀerentiating term by term with respect to x we ﬁnd ∞ ux + ivx = n=0 nan (x + iy)n−1 . ux + ivx = vy − iuy which implies ux = vy and vx = −uy . 19.271 and 2 L L Bn = (f2 (x) − f1 (x)) sin 0 nπ xdx L sinh nπH 2L −1 19. diﬀerentiating term by term with respect to y we ﬁnd ∞ uy + ivy = n=0 nan i(x + iy)n−1 .

y) = 1 sinh ∞ 8π 3 4π(x − 2) 3 u4 (x.8 u(x. 19. y) + u2 (x.1) we obtain the dersired equation.272 rule we obtain ux =ur rx + uθ θx = cos θur − ANSWERS AND SOLUTIONS sin θ uθ r uxx =uxr rx + uxθ θx sin θ sin θ = cos θurr + 2 uθ − urθ cos θ r r cos θ sin θ sin θ + − sin θur + cos θurθ − uθ − uθθ − r r r cos θ uθ uy =ur ry + uθ θy = sin θur + r uyy =uyr ry + uyθ θy cos θ cos θ = sin θurr − 2 uθ + urθ sin θ r r sin θ cos θ cos θ + cos θur + sin θurθ − uθ + uθθ r r r Substituting these equations into (21. y) = n=1 nπ nπ 14(1 − (−1)n ) sin y sinh x . y) = 4 sinh πL 2H sinh πx − sinh 2H π(x − L) 2H cos πy .9 u(x. y) = n=1 − (−1)n 2 · nπ sinh 3nπ 2 sinh sin nπ nπ x sinh y 2 2 sin 4π y 3 u3 (x. 2H . 2nπ 3 3 nπ sinh 3 19. y) where u1 (x. y) + u3 (x. y) = u1 (x. y) + u4 (x. y) = 0 ∞ u2 (x.

19. 20.13 u(x. θ) = 2π 1 = 2π 2π ∞ f (φ)dφ + 0 2π 0 n=1 ∞ rn πan r a 2π f (φ) [cos nφ cos nθ + sin nφ sin nθ] dφ 0 n f (φ) 1 + 2 n=1 cos n(θ − φ) dφ.10 u(x. t) = A0 + ∞ n=1 An e− λn x H √ cos λn y where 1 H 2 An = H A0 = 19.14 u(x. H 3πx L − 5 Y3 (H) sin . y) = 2 cosh 3y sin 3x cosh 6 − 5 cosh 10y sin 10x .11 u(x. and Bn into the right-hand side of u(r.3 u(r. 20. 2 2 19.5 (a) We have eit = cos t + i sin t and e−it = cos t − i sin t. y) = 20 Y (y) sin Y1 (H) 1 f (y)dy 0 H f (y) cos 0 πx L nπ ydy. θ) we ﬁnd 1 u(r. 20. y) = y. 19.15 u(x. θ) = π 4 + ∞ n=1 rn 1−(−1)n n2 π cos nθ + sin nθ n . The result follows by adding these two equalities and dividing by 2. θ) = 3r5 sin 5θ.4 Substituting C0 .1 u(r. 19. y) = 1 x2 − 1 y 2 − axb y + C where C is an arbitrary constant. 20.273 √ 19.12 u(x.2 u(r. cosh 20 Section 20 20. An . 2 . y) = sin (2πx)e−2πy . θ) = C0 + r2 cos 2θ. (b) This follows from the fact that 1 cos n(θ − φ) = (ein(θ−φ) + e−in(θ−φ) ).

a − 2ar cos (θ − φ) + r2 . A 20. (b) We have ∞ 1+2 n=1 r a n cos n(θ − φ) =1 + + rei(θ−φ) a − rei(θ−φ) re−i(θ−φ) a − re−i(θ−φ) r r + −i(θ−φ) =1 + −i(θ−φ) ae − r ae −r r =1 + a cos (θ − φ) − r − ai sin (θ − φ) r + a cos (θ − φ) − r + ai sin (θ − φ) r[a cos (θ − φ) − r + ai sin (θ − φ)] =1 + a2 + 2ar cos (θ − φ) + r2 r[a cos (θ − φ) − r − ai sin (θ − φ)] + a2 − 2ar cos (θ − φ) + r2 a2 − r 2 = 2 .274 ANSWERS AND SOLUTIONS r a r (c) We have |q1 | = a cos (θ − φ)2 + sin (θ − φ)2 = similar argument shows that |q2 | < 1. < 1 since 0 < r < a.6 (a) The ﬁrst sum is a convergent geometric series with ratio q1 and sum ∞ n=1 r a n r i(θ−φ) e ein(θ−φ) = a 1 − q1 rei(θ−φ) = a − rei(θ−φ) Similar argument for the second sum.

7 We have u(r.9 (a) Diﬀerentiating u(r. t) = R(r)T (t) with respect to r and t we ﬁnd utt = RT and ur = R T and urr = R T.8 u(r. Substituting these into the given PDE we ﬁnd 1 RT = c2 R T + R T r Dividing both sides by c2 RT we ﬁnd R 1R 1T = + . 0) = g(r) = R(r)T (0). n=1 (−1) 20. If λ > 0 then T − λc2 T = 0. θ) = 2 ∞ n+1 n sin nθ r n . and the LHS depends on t only.275 20. t) = 0 = R(a)T (t) =⇒ R(a) = 0 u(r. θ) = 1 2π 2π ∞ f (φ) 1 + 2 0 2π n=1 2 r a n cos n(θ − φ) dφ 1 a − r2 = f (φ) 2 dφ 2π 0 a − 2ar cos (θ − φ) + r2 f (φ) a2 − r2 2π = dφ. 2 − 2ar cos (θ − φ) + r 2 2π a 0 20. . 0) = f (r) = R(r)T (0) ut (r. they must equal to a constant λ. If λ = 0 then R + 1 R = 0 and this implies R(r) = C ln r. Using the condition r R(a) = 0 we ﬁnd C = 0 so that R(r) = 0 and hence u ≡ 0. This equation has the solution √ √ T (t) = A cos (c λt) + B sin (c λt). (b) The given boundary conditions imply u(a. 2 T c R rR Since the RHS of the above equation depends on r only.

y) (x. y) (x.y)∈∂Ω ≤u1 (x. y) = min g1 (x.y)∈∂Ω u(x. y). y) ≤ max u2 (x. y) ≤ max u(x.12 (i) u(1.10 (a) Follows from the ﬁgure and the deﬁnitions of trigonometric functions in a right triangle.13 Using the maximum principle and the hypothesis on g1 and g2 . (b) The result follows from equation (20. 20.y)∈∂Ω (x. y) = min u2 (x. ∀(x.y)∈∂Ω (x. y) ∈ Ω. y) ≤ u(x. 0) = 4 (ii) u(−1. λ < 0. y) = u(−1. y) ∈ Ω ∪ ∂Ω we have (x. (x. y) ≤ max u1 (x. 20.1). y) < max g2 (x. y) (x. Hence.276 ANSWERS AND SOLUTIONS The condition u(r.y)∈∂Ω = max g1 (x. y) > 0 for all (x. y) ∈ Ω (x. y) = max g2 (x. 1 ≤ u(x. 20. for all (x. 0) = 1 and max(x.14 We have ∆(rn cos (nθ)) = 1 ∂ n 1 ∂2 ∂2 n (r cos (nθ)) + (r cos (nθ)) + 2 2 (rn cos (nθ)) ∂r2 r ∂r r ∂θ =n(n − 1)rn−2 cos (nθ) + nrn−2 cos (nθ) − rn−2 n2 cos (nθ) = 0 Likewise. y) ≤ 3 and this implies that u(x.y)∈∂Ω (x.11 By the maximum principle we have min u(x.y)∈∂Ω (x.y)∈∂Ω 20. 0) = 3.y)∈∂Ω ≤ min g1 (x. y) = u(1.y)∈∂Ω min u1 (x. . Hence. ∆(rn sin (nθ)) = 0. 0) = −2.y)∈∂Ω ≤u2 (x. 20. 0) = f (r) implies that A = f (r) which is not possible. y).y)∈∂Ω u(x. y) (x.y)∈∂Ω But min(x.

s2 s > 0. θ) = ln 2 + 4 cos 3θ.277 20.14 −2t + e−t . Section 21 21. s = 0.15 2(e−2t + e2t ). s−2 s > 2. t ≥ 0.13 3e2t . 1 (e−s s2 21.11 (a) 0 (b) 0.6 f (t) = e(t−1) does not have a Laplace transform.7 21.12 5 s+7 + 1 s2 + 2 . t ≥ 0. s + n s 21. tn−1 e−st dt. 21. s > 0. s > 1. s .1 Convergent.8 4 s − 4 s2 + 2 . 21. t ≥ 0.10 − t n e−st −2s − e−2s ). s > 0.9 − e s + 21.4 21. 21.15 u(r. 21. s3 s > 0. 21. a 3 r 20.16 2 s−1 + 5 .2 Divergent. 21. 21. 21. θ) = 1 2 − r2 2a2 cos 2θ. e−s . 21.5 1 .16 u(r. − 5 .3 Convergent. s−3 1 s2 s > 3. s 2 21.

26 4[e3(t−5) − e−3(t−5) ]H(t − 5). et 2 21. 1 21.24 0.20 21. t ≥ 0. 21. 3 t 2 sin t. 6 21. (s−3)2 +9 2 (s−4)3 + + 5 . t ≥ 0.30 21. t ≥ 0. 3 (s−4)2 s−3 .21 e−s . s > 3. 21.28 5 e3t + 1 −2t e 20 9(t−2) 0≤t<2 . 21. 3 .278 21.18 21.27 y(t) = 2e−4t +3[H(t−1)−H(t−3)]−3[e−4(t−1) H(t−1)−e−4(t−3) H(t− 3)]. − 1 e2t .31 21.19 21. s2 +26 s > 0. e 21. t ≥ 0.29 21. t ≥ 2.22 2 sin 5t + 4e3t . − s2 s2 +4ω 2 . s−4 s > 4. 120 1 2 1 − e−t + 2 e−2t . 2 Section 22 . t5 . s > 0. s−3 1 2 1 s ANSWERS AND SOLUTIONS s > 3.25 3e3t − 3e−t . 21. t ≥ 0.32 et −e−2t . 4 21.23 5 e3t t3 .33 −t + − e−t . t ≥ 0.17 21.

2 u(x.14 u(x.4 u(x. t) = L−1 e− c x s2 +1 s =h t− x c sin t − x c .12 u(x. t) = 2e−4π t sin πx + 6e−16π t sin 2πx. nπ 1 2 23. t) = 5e−3π t sin (πx). 22.5 u(x. 22. 22. t) = sin (x − t) − H(t − x) sin (x − t). t) = [sin (x − t) − H(t − x) sin (x − t)]e−t .9 u(x. 22. 22. Section 23 23.3 u(x. 1 22.11 u(x.6 u(x. 22.8 u(x.1 (−1)n i .1 u(x. t) = 40e−t cos x . t) = t2 e−x − te−x + t. 22. t) = t − 1 x2 H t − 2 x2 .2 f (x) = + ∞ 1 n=1 nπ sin nπ 2 (einx + e−inx ). t) = L−1 − T e− s 2 √ s x c + T s . t) = e−5x e−4t H(t). y) = y(x + 1) + 1. t) = [sin (x − t) − H(t − x) sin (x − t)]et . 2 2 2 22.15 u(x.10 u(x.7 u(x.279 22. 2 22. 22. t) = L−1 e− c x s2 +1 s =H t− x c sin t − x c . . 22. t) = 3 sin πx cos 2πt. 22. t) = 2 sin x cos 3t. 22.13 u(x.

n=1 n2 (−1) e 1 2 a0 =2 −1 2 1 2 π π 2 sin nπxdx = − [cos − cos − ] = 0 π 2 2 sin nπx cos 2nπxdx = 0 an =2 −1 2 1 2 bn =2 −1 2 sin nπx sin 2nπxdx = 8(−1)n n π − 4n2 π (b) f (x) = 23.8 (a) 4 π ∞ (−1)n n 2nπix .4 f (x) = 23. n=−∞ i(1−4n2 ) e 1 a0 = 2 1 an = 2 1 bn = 2 (b) f (x) = 2 + 2 (2 − x)dx = 4 −2 2 nπ x dx = 0 2 −2 2 4(−1)n nπ (2 − x) sin x dx = 2 nπ −2 (2 − x) cos + ∞ 2(−1)n+1 i ( inπ x) e 2 . n=1 nπ 4 nπ −1 2(−1)n+1 i ( inπ x) e 2 n=−∞ nπ 4 23.3 f (x) = 23. + ∞ 2 n inx .9 an = cn + c−n = 0.280 23.6 (a) f (x) = 2 (b) f (x) = π3 + 23. We have for |n| odd bn = i inπ = |n| even bn = 0.7 (a) −1 2 n inx n=−∞ n2 (−1) e ∞ 4 n n=1 n2 (−1) cos nx. Thus.5 f (x) = sinh aπ π eix −e−ix . and for 23. 2i 1 2π ∞ (−1)n (a+in) inx e . n=−∞ (a2 +n2 ) ANSWERS AND SOLUTIONS T+ π2 3 −1 i −int n=−∞ n [e − 1]eint + + ∞ i −int n=1 n [e − 1]eint .10 Note that for any complex number z we have z + z = 2Re(z) and z − z = −2iRe(z). 23. c n + c n = an .

281 which means that an = 2Re(cn ). 23. bn = −2Im(cn ). n=−∞ 2−inπ 23. nπ (c) We have 1 bn = 0 sin nπxdx = 1 − (−1)n 1 − cos nπ = . bn = (d) We have c0 = a0 2 2 nπ 0 = 1 2 and for n ∈ N we have an − ibn = 2 i − nπ if n is odd 0 if n is even cn = 23. nπ ∞ i sin (2−inπ) inπ x e 2 .14 sin 3x = 1 (e3ix − e−3ix ).11 an = 2Re(cn ) = 23.15 e−ins 1−e−inh 2πin .12 f (x) = i 1 πn sin (nT ) and bn = 1−cos (nT ) . Likewise. Hence. . (b) We have 2 a0 = L 1 L 2 1 1 0<t<1 0 1<t<2 f (x)dx = 0 0 dx = 0 dx = 1 an = 0 cos nπxdx = sin nπ = 0.13 (a) We have f (t) = and f (t + 2) = f (t) for all t ∈ R. 2 23. nπ nπ if n is odd if n is even Hence. we have cn − cn = ibn That is ibn = −2iIm(cn ).

ˆ ˆ 24.5 1 α−iξ 2 sin ξa . 0) = f (ξ). 1 + iξ 24.6 We have ∞ F[e−x H(x)] = −∞ ∞ e−x H(x)e−iξx dx e 0 −x(1+iξ) = e−x(1+iξ) dx = − 1 + iξ ∞ = 0 1 . ξ + 1 α+iξ = 2α . 0) = f (ξ) ˆ ut (ξ.4 uyy = ξ 2 u ˆ ˆ u(ξ.282 ANSWERS AND SOLUTIONS Section 24 24. we have F 1 = F[F[e−ξ H(ξ)]] = 2πeξ H(−ξ). 0) = g (ξ).7 Using the duality property. u(ξ. ˆ 24. α2 +ξ 2 24.2 2 sin ξ if ξ = 0 ξ 2 if ξ = 0. ∂u ˆ + iξcˆ = 0 u ∂t ˆ u(ξ. L) = ˆ ˆ 24. 1 + ix .3 ∂ 2u ˆ = −c2 ξ 2 u ˆ 2 ∂t ˆ u(ξ. 0) = 0.1 ˆ f (ξ) = 24.

24.10 We will just prove the ﬁrst one. 2 24.8 We have ∞ F[f (x − α)] = −∞ f (x − α)e−iξx dx ∞ =e−iξα −∞ f (u)e−iξu du ˆ f (ξ) =e where u = x − α.9 We have ∞ −iξα ∞ F[eiαx f (x)] = −∞ eiαx f (x)e−iξx dx = −∞ ˆ eix(α−ξ f (x)e−iξx dx = f (ξ − α). − cos ξa). .11 Using the deﬁnition and integration by parts we ﬁnd ∞ F[f (x)] = −∞ f (x)e−iξx dx ∞ ∞ = f (x)e−iξx + (iξ) −∞ f (x)e−iξx dx −∞ ˆ ˆ =f (x) cos ξx − if (x) sin ξx + (iξ)f (ξ) = (iξ)f (ξ) where we used the fact that limx→∞ f (x) = 0.283 24.13 2 (1 ξ2 2 (1 iξ − cos ξ).12 24. We have F[cos (αx)f (x)] =F[ f (x)eiαx e−iαx + f (x) 2 2 1 = [F[f (x)eiαx ] + F[f (x)e−iαx ]] 2 1 ˆ ˆ = [f (x − α) + f (x + α)]. 24. 24.

t) = = (x−ct)2 4 . t)] = e− 25. u(x.5 √1 4πkt ∞ − (x−s) e 4kt 0 2 ds. γ −αt −1 −ξ2 (kt+ γ ) 4 ] e F [e 4π x2 γ −αt π e · · e− 4(kt+γ/4) 4π kt + γ/4 x2 √ = γ4kt + γe− 4kt+γ e−αt . 4πt 25.1 u(x. t) = 25. x ≥ 0.284 2 ANSWERS AND SOLUTIONS ˆ 24. 25. t) = f (x) ∗ F −1 [− |ξ| e−|ξ|y ]. t) = F −1 [u(ξ.6 We have ∞ 0 ∞ 2 e −∞ −|ξ|y iξx e dξ = −∞ e e ξy iξx dξ + 0 0 e−ξy eiξx dξ ∞ 0 1 1 eξ(y+ix) + eξ(−y+ix) y + ix −y + ix −∞ 1 1 2y = + = 2 . Section 25 1 25.2 u(x.4 u(x.3 u(x. = e−ax . 25.15 F −1 1 a+iξ x √1 e− 2 2π .14 F −1 [f (ξ)] = 24. y + ix −y + ix x + y2 = . t) =et F −1 [e−ξ t ] 1 − x2 =e−αt √ e 4t .

12 u(x.10 u(x.8 utt + (α + β)ˆt + αβ u = −c2 ξ 2 u. t) = f (x) ∗ F −1 [− |ξ| e−|ξ|y ].285 25. 2π −∞ (x − ξ)2 + y 2 25. t) = e−(x−3t) . t) = e−(x−kt) .7 u(x. y) = ∞ 1 ˆ f (ξ)e−|ξ|y eiξx dξ 2π −∞ 2y 1 = f (x) ∗ 2 2π x + y2 ∞ 1 2y = f (x) dξ. 25. 1 25.13 u(x. ˆ u ˆ ˆ 25. . t) = (x − ct)2 . t) = √1 4πkt 2 (x−s) ∞ e−s − 4kt −∞ 2 ds.9 u(x. 25. 25.11 u(x.